Geotechnical instrumentation for both routine monitoring and research was conducted during construction of the new Harvard Square Station, part of expansion of a rapid transit line in metropolitan Boston. The cut‐and‐cover structure was designed with slurry walls (cast‐in‐place concrete diaphragm walls) that were used both for initial excavation support and as final structural walls. The purpose of the study was to develop a better understanding of slurry walls systems that will benefit future design and construction of slurry walls as permanent structures. In conjunction with analytical studies, the field measurements provided better understanding of slurry wall behavior both during construction and the long term. The field measurements summarized in this report focused on lateral movements, bending in the wall as measured by strain gauges, and settlement of the wall and adjacent structures. Field measurements showed movements to be small, and bending in the stiff wall to be less than design values. Design recommendations are given.

The uncertainties that are encountered in prediction of offshore foundation performance are reviewed. An outline of the first‐order, second‐moment method is given. This method is used to calculate the effect of uncertainties about the input on the predicted performance and reliability. Reliability analyses for an artificial island, gravity platforms on sand and on clay, and a jacket platform on piles are summarized. These examples demonstrate the use of reliability analysis in the design‐decision making process.

Both deformational response and stress distribution in an anisotropic rock mass are examined. Joint imposed rock mass anisotropy is considered in relation to joint and rock material shear stiffness. Two typical rock mass deformational characteristics are investigated by considering component moduli representing a firm and then soft rock mass. These component moduli are combined to give composite moduli and resultant directional rock mass Young's modulus. The resultant Young's modulus is defined first from classical infinitesimal elastic theory and then considering specific deformations. There are major differences between the two definitions involving shear moduli and Poisson's ratio. For an anisotropic rock mass, the induced horizontal stress derived from a rock bolt produces an armoring effect that is so often observed in practice. Rock bolt effects on rock mass anisotropy and deformability are investigated by considering joint shear characteristics.

A probabilistic model for one‐dimensional consolidation analysis of a clay deposit is formulated based on conventional theory for prediction of time history and magnitude of consolidation settlements. The model accounts for uncertainties in the loading as well as in the deformation properties of the clay. The model is updated through in‐service settlement measurements, and measurement uncertainties are considered. The practical application of the model is illustrated by presentation of calculations for an example foundation. Probability calculations are performed by first‐order reliability methods.

A simple and convenient semi‐analytical procedure is presented to calculate the factor of safety of embankments constructed on soft clay. Stability numbers $N_1$ and $N_2$ are developed for the normalized foundation strength and normalized embankment strength, respectively. The factor of safety (in terms of overall moment equilibrium) can be computed as the sum of two components, each equal to the product of the respective stability number and the corresponding normalized strength. The relative importance of the foundation strength and the embankment strength in contributing to the factor of safety can also be assessed during the computation. The proposed method assumes short term undrained response of the soft clay foundation. It can deal with cases where the undrained shear strength of the soft clay varies with depth, and where the embankment possesses both cohesion and internal friction. The application of the method is illustrated by several examples and the results are compared with those obtained from a computer analysis.

This paper is the culmination of a three‐phase research effort to estimate the preconsolidation pressure $p_c$ of in situ soils using a conventional Menard pressuremeter modified for acoustic emission (AE) monitoring. The first two phases involved determination of the feasibility of the AE technique using controlled laboratory tests on both granular and cohesive soils that were prestressed to known $p_c$ values. These results were previously reported and have shown that the AE data manifested itself as a bilinear response curve that “broke” at pressures approximately $\pm 10\%$ of the known $p_c$ of the soil. This paper reports on the complementary field phase in which six field sites were tested. The in situ soils were all finegrained residual soils derived from weathering of the parent bedrock. All had some degree of preconsolidation. In general, the results of this investigation have shown that the AE data again produced bilinear response curves that “break” at a particular value of applied horizontal pressure. This value of horizontal pressure was found to compare reasonably well [i.e., within $\pm 143.6$ kPa ($\pm 1.5$ tsf) with the horizontal component of $p_c$ (i.e., $p_{\mathrm{cH}}$) determined from laboratory tests. These latter test results were determined from conventional consolidation tests performed on Shelby tube samples of the residual soils, which were trimmed vertically so as to yield their horizontal compression properties. Presently, the described “acoustic pressuremeter” appears to be ready for general use by the profession to assess its advantages and/or disadvantages.

A simplified method for finding rotational boundary stiffnesses, representing the foundations of a structure and its surrounding soil, is developed. This technique is applied to the Meloland Road overpass, for which records were obtained during the 1979 Imperial Valley earthquake. A system identification method was previously used to find the predominant modes of the structure. A very simple, finite element 3‐D beam model for the bridge is constructed, and the rotational stiffnesses are used to model the effect of the foundations. A dynamic analysis is performed, in which the recorded translational earthquake accelerations are used as input excitation to the structure, and the predominant modal frequencies of the model are shown to compare well with the results obtained through system identification, even though simplifying assumptions were made. A sensitivity analysis is also performed on the rotational boundary springs to study how much they affect the dynamic response of the structure. In further studies, these procedures could be used to gain more insight into the dynamic effects of soil‐structure interaction and should prove to be particularly useful for preliminary designs of structures.