Reinforced soil technologies have been extensively used during the past few decades in the construction of embankments and retaining walls. A large variety of reinforcing materials emerged and have been developed for construction purposes, including: metal strips, bar mats, geotextile sheets, geogrids, etc, Fullscale experiments and centrifugal and reduced‐scale laboratory model studies have been performed to provide a rational basis for design methods of reinforcing systems using quasi‐inextensible steel strips and bar mats. However, laboratory model studies on the behavior of reinforced soil walls with more extensible reinforcements, such as geotextiles and geogrids, have been rather limited. These studies have encountered fundamental difficulties pertaining to similitude requirements and specifically appropriate modeling of the reinforcing materials, as well as to the proper instrumentation of extensible reinforcing elements to measure the mobilized tension forces. This paper describes the results of a laboratory model study on the performance, behavior and failure mechanisms of reinforced soil‐retaining walls using different reinforcing materials, namely: woven polyester geotextile strips, plastic grids, and nonwoven geotextile strips. The model walls were instrumented to obtain measurement of stresses in the reinforcements, displacements at different points along the reinforcements, displacements of the facing, and to identify, using colored sand, the failure surface in the soil. Experimental results indicate that the confinement of the reinforcement has a major effect on the structure performance. The paper also outlines practical conclusions that can be drawn for design purposes.

Final results of an extensive laboratory research program on “critical” downstream filters in embankment dams with impervious sections are presented. The investigation was carried out at the Soil Mechanics Laboratory, Midwest National Technical Center, Soil Conservation Service, United States Department of Agriculture, Lincoln, Nebraska. A new laboratory test, called the “no erosion filter” (NEF) test, was found to have substantial advantages over the slot and slurry tests used in the earlier part of the research program for silts and clays. It gave essentially the same results for silty and clayey sands with low fines content as was obtained with conventional filter tests on uniform sands. A wide range of different fine silts and clays and clayey and silty sands of different geologic origins were tested. The results of the investigation confirm conclusively that sand filters containing appropriate quantities of fine sand will reliably control and seal concentrated leaks through the impervious sections of embankment dams. The investigations show that for most fine silts and clays, a downstream sand filter with $D_{\mathrm{l5}}⩽0.5\mathrm{mm}$ is conservative and that broadly graded soils, such as those from glacial moraines, need a relatively fine filter.

If a tunnel crosses a geological fault which is considered to be active, the possibility of a displacement in the fault must be accounted for in the tunnel design. When the tunnel is embedded in soil, the assessment of the effect of the fault movement is not easy to assess. As a guide to analysis, a series of correctly scaled model experiments in a centrifuge is undertaken. The deflections and stresses induced in a tunnel crossing a fault caused by: (a) Fault displacement; and (b) differential earthquake motions across the fault are quantified by a series of centrifuge tests on a finite length model tunnel. The centrifuge results are used to calibrate a one‐dimensional finite element model of the tunnel for soil‐tunnel interaction effects. The numerical model is then used to predict the response of an essentially infinite length tunnel for design purposes. Bending movements, displacements and shears are displayed. Surprisingly small changes in the bending movements from the finite‐length to the infinite tunnel case are obtained.

The cyclic behavior of normally consolidated clays is investigated by conducting a series of cyclic simple shear tests on one‐dimensionally and isotropically consolidated kaolinite samples. The response of two types of specimens are evaluated and compared based on tests performed with different cyclic shear stress amplitudes and frequencies. The cyclic stress‐strain‐pore pressure and cyclic shear strength properties of the clay samples are studied and a simple procedure is suggested to define cyclic yield strength. The results also indicate that there is a threshold cyclic shear stress level below which no excess pore pressure will develop. An attempt is made to develop an empirical procedure to estimate the response of a soil element subjected to cyclic shear stresses for a given number of cycles.

Pile foundation studies were conducted on four types of steel piles driven through estuarine deposited soils and into a highly variable glacial deposit to total depths of 33–48 m (110–156 ft). Twenty piles were subjected to dynamic monitoring during initial driving and restriking. For all the monitored piles, static loading tests were carried out to failure and the results compared to ultimate resistances determined in a CAPWAP analysis. With two exceptions, the driving was generally very easy above a depth of 46 m (150 ft). Restriking at different times after driving showed that the penetration resistance increased due to the development of soil set‐up occurring within the first week after initial driving and that the final pile capacities vary considerably and randomly across the site. The CAPWAP‐determined pile capacities at restriking agreed well with the results of the static loading tests (when the latter could be clearly defined and the hammer had been able to move the pile in the restriking). When the capacity could not be defined in the static test, the CAPWAP determined ultimate resistance was as usable as the conventional methods for determining the load limit of the static test.

TO compute the degree of saturation, S, one must have the results from tests for water content, wet unit weight, and specific gravity. Each has inherent variability, which results in a variability in S. This variability results in a percentage of S plotting above the zero air voids line (ZAVL). After first showing that Gaussian normal distributions apply to the test results, the writer proposes typical low, average, and high values for test and soil plus test variability and then uses both numerical and analytical methods to predict the percent of S above the ZAVL. He then presents the density test results from three large cohesive fills to support the analytical study. Knowing the percent of tests above the ZAVL may provide the engineer with an additional quality check of a cohesive fill. Arbitrarily rejecting all tests above the ZAVL will distort the perceived character of the fill, which could result in action detrimental to the engineering behavior of the fill and might provide the basis for a claim for extra compensation by the contractor.

Data are presented for thirty-three geotechnical centrifuge models of reinforced soil retaining walls in which the effects on wall behavior of foundation and retained fill are studied. The results indicate that the retained fill has little influence, but that the effect of the foundation is considerable, although it is usually underestimated in practice.