A review is made of the hydraulic conductivity and other hydraulic parameters of municipal waste. The hydraulic conductivity can be assessed indirectly from measured field parameters and water balance. A test well penetrating about 100 ft of refuse was installed and pumped for about one day at 20 gpm and 2‐1/2 days at 12 gpm. Drawdowns were measured at three observation wells and the pumped well. Difficulties were encountered during drilling and subsequent monitoring of leachate levels. The hydrogeologic parameters were computed using conventional hydrogeologic analysis. The results of a pumping test of leachate from a municipal landfill are presented. Based on the results of the pumping test, it is concluded that hydraulic conductivity of municipal refuse is about ${10}^{-3}\mathrm{cm}/s\text{.}$ It is also concluded that pumping of leachate from a municipal landfill is feasible for control of leachate release to ground water.

Two different models for reliability analyses of foundation failure problems are presented, one where only average soil‐strength properties are modeled, and one where the soil strength is represented more accurately as a random field. Two example cases are presented, namely static pull‐out of a vertical pile and horizontal sliding of a gravity‐base structure, both installed in clay. Uncertainties in environmental loading as well as in soil‐strength properties are modeled. The spatial fluctuation structure of the soil strength is estimated by means of a kriging technique. Reliability calculations for the two example foundations are performed by first‐order reliability methods. For each example foundation the results of the two different types of reliability analyses are compared, and the effect of representing the soil strength as a random field is assessed.

Eleven one‐tenth scale tubular steel pile tests are conducted to establish the behavior of pinned‐head friction piles in loose sand, fixed‐head friction piles in both loose and compacted sand, and a fixed‐head bearing pile in loose sand. The piles are instrumented with electrical resistance strain gauges and the vertical and horizontal pile movements at grade were measured with displacement transducers. By differentiation and integration of pile strain functions, obtained by curve‐fitting the measured strain data, the soil resistance and displacement relationships are developed in the form of modified Ramberg‐Osgood expressions. The characteristic soil parameters, describing the skin friction and vertical displacement f‐z curves, end‐bearing and settlement p‐z curve, and lateral resistance and displacement p‐y curves, are used in finite element studies that idealized the nonlinear soil behavior by isolated Winkler springs. The analytical results for the pile strains along the embedment length and displacements at the soil surface of a particular pile test correlate reasonably well with the measured values, considering the sensitivity of techniques used to establish the soil responses.

The various formation factors responsible for differences in behavior between residual soils and sedimentary (transported) soils are described. The extent to which classical soil mechanics concepts derived from the study of sedimentary soils are applicable to residual soils is examined and discussed. It is shown that residual soils can be wrongly evaluated as problem soils simply because some aspects of their behavior do not conform to that of sedimentary soil. The relative importance of composition and structure in influencing residual soil behavior is examined by carrying out consolidation and triaxial tests on three residual soils, namely, a silt, a tropical red clay, and an andosol (volcanic ash soil). The need for an empirical or theoretical framework applicable to residual soils, in place of the stress history framework used with sedimentary soils, is discussed.

One of the most important dynamic properties required in the design of machine foundations is the stiffness or spring constant of the supporting soil. For a layered soil system, the stiffness obtained from an idealization of soils underneath as springs in series gives the same value of stiffness regardless of the location and extent of individual soil layers with respect to the base of the foundation. This paper aims to develop the importance of the relative positioning of soil layers and their thickness beneath the foundation. A simple and approximate procedure called the weighted average method has been proposed to obtain the equivalent stiffness of a layered soil system knowing the individual values of the layers, their relative position with respect to foundation base, and their thicknesses. The theoretically estimated values from the weighted average method are compared with those obtained by conducting field vibration tests using a square footing over different two‐ and three‐layered systems and are found to be very good. The tests were conducted over a range of static and dynamic loads using three different materials. The results are also compared with the existing methods available in the literature.

Classification systems define rock mass zones that represent the potential behavior of the rock mass when subjected to load changes. Rock mass characterization, as opposed to rock mass classification, is described by an example of the right‐abutment ridge, Thomson Dam, Australia. Geological investigations demarcate the discontinuities into sets covering specific zones or domains. Samples from these rock zones are retrieved and indexed and the zones are then bounded and classified. From selected sites, discontinuity samples are retrieved and laboratory‐tested for strength and stiffness. Rock material moduli and joint stiffnesses are combined using two simple models representing the constraints of equilibrium and compatibility, and produce the predicted rock mass' anisotropic moduli. Measured in situ rock mass moduli are compared with predicted moduli. Of the various discontinuity types, the bedding plane thrust faults have the lowest strengths and greatest deformational potential. Fault gouge and its association with clay content control the friction angle and deformability potential.

A slope‐stability failure occurred in a 15 acre hazardous‐waste landfill (90 ft high) in which lateral displacements of up to 35 ft and vertical settlements of up to 14 ft were measured. Failure developed by sliding along interfaces within the composite, multilayered geosynthetic‐compacted clay liner system beneath the waste fill. The testing, analyses, and related studies made to determine the cause of the failure are the subject of this and a companion paper (Seed et al. 1990). The present paper presents details of a direct shear and pullout testing program undertaken to determine liner‐system‐interface shear‐strength characteristics. The interfaces between the various geosynthetics, and between these materials and the compacted clay in the liner system, are characterized by low frictional resistance, with values of interface‐friction angle as low as 8° for some combinations. The most critical interfaces were determined to be those between high‐density polyethylene (HDPE) geomembrane and geotextile, HDPE geomembrane and geonet, and HDPE geomembrane and saturated compacted clay. Representative values of interface shear‐strength parameters were obtained for use in the stability analyses described in the companion paper. The variations in measured strength parameters for the different interfaces in the liner system indicate the desirability of conducting similar test programs for proposed new facilities to establish design parameters.

Analyses were made to determine the cause of a stability failure of a 90 ft high, 15 acre hazardous waste landfill in which lateral displacements of up to 35 ft and vertical settlements of up to 14 ft were measured. The failure developed by sliding along interfaces within the composite geosynthetic‐compacted‐clay liner system beneath the waste fill. The shear resistances of the different interfaces in the liner system were determined by direct shear and pullout tests as described in a companion paper (Mitchell et al. 1990). Conventional two‐dimensional (2D) stability analyses of representative cross sections and three‐dimensional (3D) analyses of the overall waste fill and liner configuration are described. Each type of analysis was applied to two cases: (1) The “Probable Minimum Clay/Liner Wetting Case,” in which shear along a wetted HDPE liner/compacted clay interface was assumed to occur only over a small area of the base; and (2) the “Full‐Base‐Wetting Case,” in which the HDPE liner/compacted clay liner interface was assumed to have become “wetted” over the full central base of the fill basin. The 2D stability analyses gave factors of safety of 1.2‐1.25 and 1.1‐1.15 for the minimum wetting case and the full base wetting case, respectively, while the 3D analyses yielded values of 1.08 and 1.01 for these two cases. Uncertainties in the strength parameters and analysis methods lead to a best estimate of the computed factor of safety at the time of failure of 0.85‐1.25.This provides good agreement with the observed field performance, and suggests that the techniques used to evaluate liner‐interface shear strengths and to perform overall stability analyses may be appropriate for future evaluation of other, similar lined waste‐repository fills.

Settlement of and load distribution in pile groups under vertical and centered static loads are analyzed. The analytical method incorporates the fact that shaft-load over point-load ratios vary throughout loading, thus coupling the effects of shaft loads and point loads on group settlement calculations. Other available methods uncouple shaft loads and point loads using simplifying assumptions that are not always in agreement with observed pile group behavior. It is assumed that the shaft load of piles in groups increases monotonically with loading according to a nonlinear relationship. This can be defined from results of a reference pile load test with point-load measurement, for instance. It is also assumed that the ultimate shaft load develops before he ultimate point load. These assumptions, which are in general agreement with observed behavior, are used along with settlement influence factors and settlement interaction factors numerically generated for a semi-infinite, homogeneous, linear, elastic, isotropic half space. The result is a set of nonlinear simultaneous algebraic equations that can be solved using an iterative numerical procedure to yield the group settlement, as well as point loads and shaft loads for each pile in the group. An example is provided to illustrate how the method compares with two other methods reported in the literature. The proposed method has the potential to accurately simulate the nonlinear load versus settlement behavior.

A study is presented on the resistance to liquefaction of sands with different gradations but identical mean grain size and mineralogy. Test specimens are reconstituted using a newly developed slurry deposition technique that yields not only homogeneous unsegregated specimens, but also simulates closely fluvial and hydraulic fill deposition processes. It is shown that with improved gradation, the cyclic strength at a given relative density increases in the range of lower relative densities but decreases in the range of higher relative density. In the range of loose relative densities, poorly graded sand develops strain as a consequence of contractive deformation in contrast to cyclic mobility for sands that are better graded. Thus, gradation may control the occurrence of contractive deformation and, hence, possible flow failure at low relative densities.

This paper shows how the safety factor against a flexible structure overturning on compressible ground depends not only on the weights and lever arms involved but also on the height of the weights above the ground. A rigid structure on rigid ground (without piles) resists overturning under applied loads by the restoring moment of the structure's weight acting about an axis of rotation at the bearing edge. In contrast a top-heavy structure on soft ground can overturn like a boat without any applied loads or warning, because any sway of the elevated weight creates a over-turning moment that causes further rotation and sway on the soft ground, and so on uncontrollably. These two conditions represent the extremes of a single pattern of structure-soil interaction. Worked examples illustrate reductions of an apparent safety factor of about 20% for a crane and an offshore jack-up platform on soft ground.