Some compacted soils wetted under certain conditions of compaction and loading may exhibit additional settlement (collapse). To investigate the effect of the soil type, compaction water content, initial dry unit weight and applied pressure at wetting on collapse potential, 138 single‐oedometer tests were conducted on eight different soils. The results indicate that well‐graded soils tend to collapse more than poorly graded ones under similar conditions. Furthermore, collapse potential decreases with an increase in (1) The difference between the sand and clay percentages; (2) compaction water content; and (3) initial dry unit weight, while it increases with pressure at wetting. A multiple regression analysis was performed on the laboratory results to provide a collapse‐predictive model that was substantiated by data reported by several researchers. A chart is presented to estimate (1) The collapse potential of soils under various conditions; and (2) the critical water content beyond which no collapse is expected. Additionally, a means by which collapse‐susceptible soils can be classified is provided.

Soil nailing as a method of slope stabilization and repair finds widespread application despite the absence of a clear and unified design approach. The ability of soil nails to increase the shear strength of soil by acting under combined loading, namely, shear and tension, is one of the more controversial aspects of design. For any structural member there is a relationship between the maximum bending moment and axial force, where the bending moment is in turn a function of the shear force in the member. The magnitude of this shear force can be deduced by either an elastic or a plastic model of the soil‐nail interaction. The analysis presented in the paper is for both grouted and ungrouted nails, and design equations are given. The finding is that for typical soil‐nail diameters only a small proportion of the maximum shear strength of the nail can be mobilized. Expressed in terms of the improvement in soil shear resistance, the benefit from using shear force is found always to be modest when compared with that due to the axial force in the bar

The effects of multiple modes on Rayleigh wave dispersion are discussed to reduce the ambiguity of uniqueness of shear wave velocity $\left(V_s\right)$ profiles estimated by the surface wave method. Based on a review of previous studies, dispersion curves of multiple‐mode Rayleigh waves induced by harmonic vertical point loading are derived for both vertical and horizontal particle motions. Also presented is the variation with frequency of the amplitude ratio between horizontal and vertical particle motions. Numerical studies indicate that a stiff soil layer overlying a softer soil layer induces a higher mode or multiple modes, leading to an inversely dispersive characteristic. Consideration of the effects of higher modes is strongly recommended in the inverse process when the observed data show an inversely dispersive trend. The ambiguity of uniqueness of the inverted soil profiles may be reduced by using either the dispersion data of horizontal motion or the amplitude ratio of particle motions in addition to the dispersion data of vertical motion.

A possible use of short‐period microtremors is examined for the determination of shallow shear wave velocity $\left(V_s\right)$ profiles. Presented for this purpose is a hybrid version of Rayleigh wave methods, in which vertical ground‐surface motions of either microtremors or forced vibrations are measured using an expanding circular array of six vertical sensors. Based on high‐resolution frequency‐wave number spectrum analysis of these data, Rayleigh wave dispersion curves can be determined. Microtremors cover the characteristics in a low‐frequency range, while forced vibrations cover them in a high‐frequency range. An inverse process that takes into account the effects of multiple Rayleigh modes on the measured dispersion data enables one to estimate $V_s$ profiles from the ground surface to depths more than 50 m. The $V_s$ profiles inferred by the proposed method are compared with those obtained by the conventional down‐hole method. The $V_s$ profiles from the two methods are in fairly good agreement, indicating that the use of short‐period microtremors is promising as an economical means of $V_s$ profiling.

The results of a generalized variational slope stability analysis, which is rigorous in the framework of limit equilibrium, are presented. To alleviate possible intimidation due to the abstract nature of the variational analysis, interpretation is provided showing this analysis to be equivalent, in principle, to traditional rigorous limit equilibrium methods. The variational approach, however, is more efficient in the sense that no guessing of a force‐related function is needed to obtain a statically determinate problem, nor is it necessary to assess the results for static reasonableness. Consequently, its application is simpler than other rigorous methods. A simple yet efficient numerical scheme, requiring the solution of only three simultaneous nonlinear equations, is illustrated. All the algorithms necessary to assemble a computer program are explicitly given. Several complex problems are analyzed, comparing the results with the predictions of the Morgenstern‐Price and Janbu methods. This comparative study indicates the applicability of the variational analysis, suggesting it is on par with other rigorous methods.

In December 1985, construction began on the offshore portion of the Renton Effluent Transfer System. Under the auspices of the Municipality of Metropolitan Seattle (Metro), treated effluent is pumped to Duwamish Head where the flow is separated into two 1.63‐m‐diameter steel pipes that extend about 3,050 m offshore into Puget Sound. A 155‐m long diffuser section at the end of each pipe rests in about 185 m of water. A submarine slide occurred in Puget Sound during the dredging operation for the construction of the offshore portion of the pipeline. The slide occurred during low tide and involved $400\text{,}000m^3$ of medium‐dense sands to a thickness of up to 12 m. The failure and an evaluation of the contributing causes are described in this paper. The failure is attributed to a collapse mechanism in the soil causing liquefaction as a result of low tidal conditions combined with details of the dredging operations.

The hydraulic conductivity of three $2.9m^2$ (32 sq ft) geosynthetic clay liners (GCLs) was measured. Tests were performed on individual sheets of the GCLs, on overlapped pieces of GCLs, and on composite liners consisting of a punctured geomembrane overlying a GCL. Hydraulic conductivities of two of the GCLs were in the range of ${10}^{-10}\mathrm{to}{10}^{-8}\mathrm{cm}/s\text{.}$ No flow was measured through the third GCL, but the conductivity was obviously very low $(<<{10}^{-7}\mathrm{cm}/s\text{.}$ The hydraulic conductivities of overlapped GCLs were about the same as those of the control samples with no overlap; an effective hydraulic seal developed along the overlaps in all of the materials tested. Performance of the punctured geomembrane‐GCL composites varied—performance was best when the punctured geomembrane was placed directly against bentonite and no geotextile separated the punctured geomembrane from the bentonite. For those GCLs with geotextiles on both sides, problems with migration of bentonite into the underlying drainage layer were encountered when inadequate filtration was provided. However, with a suitable filtration layer separating the drainage layer from the GCL, problems with migration of bentonite were eliminated.

The stress‐strain‐srength responses of compressible Chicago glacial clays are presented.These clays are lightly overconsolidated supra‐ or subglacial tills. Fixed piston and Shelby tube samples from various locations in the Chicago area were subjected to an extensive laboratory testing program, including index property, consolidation, direct shear, and $K_0$‐consolidated, and undrained triaxial compression and extension tests. Results of the experimental program indicated the strengths of tills can be normalized when proper consideration is given to variations in natural water content. The different undrained strengths in compression and extension indicate that the clays exhibit stress‐induced anisotropy. Initial undrained elastic moduli of the clays vary systematically with the overconsolidation ratio. Comparisons are made between strengths computed from relations based on normalized undrained strengths determined in the laboratory and those based on in‐situ tests made at several locations. These data suggest that field vane strengths represent a conservative index of undrained shear strength for these Chicago clays.

Increased usage of drilled shafts to carry high‐foundation loads has resulted in the common practice to socket the shafts into bedrock to transmit these loads. However, end‐bearing uncertainties have lead various designers to rely solely upon skin friction within the sockets to transmit the loads. This paper reviews several design techniques/correlations to estimate rock/shaft friction values from lab tests. These methods rely upon: (1) Unconfined compression $\left(q_1\right)$ and (2) split tensile $\left(q_1\right)+q_u$ tests on recovered rock cores to estimate the rock shear transfer values. A database of 14 field‐load tests compares the measured versus estimated rock shear values for the various methods, and suggests that a two‐parameter model $(q_u+q_t$ tests) provides the best estimate or socket friction. A small‐scale field “pull‐out” test on grout anchors is suggested as an alternative method for estimating rock/shaft friction values.