La Villita is a 60 m high earth and rockfill dam located 350 km southwest of Mexico City, Mexico. Since its construction the dam has been subjected to six major earthquakes, of which five are believed to be responsible for significant localized deformation in the dam. The dynamic response of the dam to these shaking events was recorded by a strong‐motion instrument located near the crest on a portion of the dam, which appears to move as a sliding block. In addition, permanent displacements after each earthquake were recorded and documented. In this investigation, it is found that the strong‐motion records not only indicate localized deformations (partial incremental failure) but also reveal the magnitude of the associated yield acceleration. The present analysis may lead to an increased understanding of the response of structures where stick‐slip‐type deformations take place.

Field trials were initiated to investigate the feasibility of minimizing secondary as well as enforcing primary compression of peaty clays by dynamic replacement and mixing (DRM) with sand charges using conventional heavy tamping plant. Laboratory model tests were also conducted to amplify the field findings. The investigations have shown that DRM can transform in situ peaty clay deposits into an upper sand raft with pockets of peaty sand underlain by a fairly uniform layer of mixed sand and peat, where both sublayers have relatively insignificant secondary compression characteristics. Furthermore, for practical purposes, unique relationships may be established between degree of improvement of peaty clay after treatment (in terms of primary and secondary compression properties) and sand‐charge‐to‐compressible‐soil ratio s. Similar relations may be formulated between s and an operational factor il, incorporating initial soil consistency and energy‐application parameters used in standard practice. By means of these characteristic relationships, overall operational requirements of a tamping scheme to achieve a desired degree of ground improvement may, in principle, be determined prior to field application.

This paper describes the results of a numerical simulation of the observed field behavior of infiltration into an unsaturated clay. In particular, the autumnal recharge of 1984 is considered. A numerical solution of Richards's pressure‐based formulation is presented. A weighted residual finite element approach is adopted to achieve spatial discretization of the moisture flow domain. A fully implicit backward difference time‐stepping algorithm is employed to accommodate the transient nature of the problem. The application of the algorithm to the simulation of the problem at hand is described. The initial moisture content distribution is established from field‐measured data. Boundary conditions are determined from an interpretation of meteorological data. Material parameter identification based on experimental work on undisturbed clay samples is detailed. Comparisons of numerical results with field values collected by British Gas are given. Good correlation is obtained. The numerical simulation is capable of representing the observed field behavior, over the depth of summer moisture deficit penetration.

A methodology is presented for predicting postconstruction reliability of full‐scale soil liners given a set of direct and indirect (compaction moisture content and dry density) permeability measurements from a test fill. The permeability is considered as a spatially correlated, lognormally distributed random variable. Six systems‐reliability measures are defined to characterize a soil liner represented by a system of blocks and lifts. The reliability measures are weakest link, averaging and parallel systems of blocks and lifts. Geostatistics is used to transfer test‐fill information to full‐scale liner permeability/reliability estimates. A relationship is developed for the number of permeability measurements necessary to monitor construction of the full‐scale liner, the target average permeability, and the specified reliability limit. The methodology is illustrated with a numerical example. It is shown how direct permeability measurements and surrogate tests commonly performed during soil‐liner construction can be combined. Reliability bounds, based on effective hydraulic permeability, can be obtained by considering various simple averaging schemes over a multilift soil liner.

The infiltration of ponded water beneath a footing on collapsible soil is monitored. Soil suction, water content, and footing settlement measurements are made. Laboratory infiltration tests are used to predict the extent of wetting in the field. Close agreement between observed and predicted extent of wetting is obtained. The in situ values of soil suction are quite high, corresponding to a pF value of about 5.26 (1,820 m of water) and degree of saturation of about 15%. In response to ponding, the soil suction drops by a factor of about 10 behind the wetted front, but remains high, with pF 4‐4.5 (100‐316 m of water) and degree of saturation averaging about 50%. Laboratory collapse tests show that partial wetting produces only partial collapse, with 50% saturation producing about 85% of full collapse strain for the soil tested. Partial wetting is considered in making footing settlement predictions, and the predicted settlement exceeds the observed settlement by only about 12%. It appears reasonable to expect close agreement, as in this case, when the extent of wetting is known in advance or well‐predicted.

TWO subdivisions in southern California constructed on compacted‐sandy‐clay fills up to 100 ft thick experienced settlement damage several years after construction. These fills were designed and constructed using the same procedures that had been used successfully in the area for many earlier developments. Field investigation showed that the water content in the fill had increased considerably after construction. This increase in water content could have been caused by rainfall, by irrigation of lawns and plants, and by infiltration of ground water into the fill from the underlying natural formations. Laboratory tests on compacted samples were used to measure compression due to wetting, and a simple analytical technique was used to calculate the magnitudes of the settlements that would result from wetting. The calculated settlements effectively bounded the measured field settlements.

Concentrated organic chemicals have been shown to cause large increases in the hydraulic conductivity of compacted clay. Mechanical and chemical methods of stabilizing four different types of compacted clay against chemical attack are investigated. Mechanical stabilization using a large compactive effort (modified Proctor compaction) or application of a compressive stress ≥ 10 psi (70 kPa) is found to render a compacted clay invulnerable to attack by concentrated organic chemicals under laboratory‐test conditions. Attapulgite, a clay mineral having little electrical charge, was found to be relatively unaffected (compared to more common clay minerals such as kaolinite, illite, and smectite) by concentrated organic chemicals. Addition of approximately 7% (by weight) of lime, portland cement, or lime plus sodium silicate greatly improved the ability of compacted clay to resist attack by concentrated organic chemicals; in some cases the amended soils were less permeable to concentrated organic chemicals than the unamended soils were to water.

The effect of base projection on the active earth pressure, a classical problem in the calculation of earth pressure forces on retaining walls, is presented. Four conclusions may be drawn. The first is that the failure wedge is bounded by two surfaces, one of them intersecting into the soil and the ground surface and the other one intersecting the wall. The second is that for the case of level backfill and no wall friction, the Rankine solution may be always applied on the vertical plane connecting the heel with the ground surface, regardless of the base projections. The third is that the inclination of the resultant earth pressure force will always be greater than the slope of the backfill and less than the angle of wall friction. Finally, the fourth is that the magnitude of earth pressure force will always be greater than that given by Coulomb solution for a wall without base projection and less than the force given by a Rankine analysis on the vertical plane through the heel with no wall friction. These conclusions are made in the case of wall interference with the shear zone. If there is not interference then the Rankine solution is applicable.

Deep natural deposits of unstable or excessively compressible soils can be stabilized by installation of sand drains. Sand drains installed in highly compressible soils provide higher rates of consolidation and, consequently, attain higher shear strengths during periods of loading. To help design these sand drains this technical note presents nondimensionalized equations that were developed for the design of sand-drain spacing for time-dependent loading.

A procedure is described for evaluating shear modulus depth profiles of uniformly deposited soil strata. The method combines steady-state vibration techniques for subsoil exploration and results from the analytical treatment of surface wave propagation problems in an elastic half-space with depth-depending stiffness. The model proposed covers a wide range of real soil deposits. The shear modulus variation with depth is obtained from the measured dispersion curve by assigning the surface wave velocity to the soil conditions at a depth of 0.3 times the wavelength.