Recent years have seen rapidly growing research into applied probability and increased interest in applications to geotechnical engineering practice. Unfortunately, probability still remains a mystery to many engineers, partly because of a language barrier and partly from lack of examples showing how the methodology can be used in the decision‐making process. The following types of applications are described and illustrated in general terms: separating systematic and random errors when evaluating uncertainty in the stability of slopes; safety factors in connection with analysis of liquefaction; optimizing the design of an embankment in the face of uncertainty concerning stability; risk evaluation for an industrial facility built over potentially liquefiable soils; and risk evaluation for earth dams. Even when a precise quantification of probability of failure is not possible, systematic formulation of an analysis aids greatly in understanding the major sources of risk and thus points the way for costeffective remedial measures. Analyses of reliability and risk are potentially most valuable during the early stages of a project in guiding the decision as to whether or not to proceed and in helping to establish design criteria.

It is shown in this paper that in a highly overconsolidated undis turbed clay, due to natural unloading, most of the strain was recovered in the vertical direction, but not in the horizontal direction. Consolidation tests were performed on both undisturbed and remolded specimens, with the specimens oriented as in conventional tests (V‐test) and at right angle (H‐test). Estimated precompression stress for V‐tests was considered inadequate to define the max imum past consolidation stress. Estimated precompression stress from H‐tests was comparatively higher, and gave a more realistic estimate for maximum past consolidation stress. However, when the soil was allowed to swell at a low stress, the H‐tests also behaved similar to the V‐tests. It is postulated, that, the lesser the degree of strain recovery since the unloading of the soil, the better is the estimate for precompression stress. Electron micrographs showed no pre ferred particle orientation for undisturbed specimens in the natural state and after loading to 26 tsf (2.6 MPa). For remolded samples, conventional tests yielded accurate values of maximum precompression stress.

Various statistical models are proposed for estimating the quality of rock masses. These models are then employed to relate RQD with the av erage number of discontinuity or average intact lengths. It is observed that the standard deviation of the intact lengths plays an additional important role in the rock quality classification. Increase in the standard deviation improves the rock quality. The proposed models are verified with the limited amount of data that are available in the rock mechanics literature. The models as a whole cover almost all of the feasible region in the RQD versus average number of discon tinuity domain.

A technique for the analysis of periodically layered soil deposits is developed. This technique requires minimal data preparation and is compu tationally very efficient. The technique is used to determine the deformations of a periodically layered half space subjected to a uniform circular applied pres sure. Consideration is given to the center settlement, edge settlement and hor izontal displacement at the edge. Results are presented for a wide range of geometries and soil conditions. The validity of two approximate approaches which represent the deposit as a single homogeneous layer are then evaluated. It is shown that simple approaches will suffice for a wide range of conditions, and parametric solutions are provided for certain situations where simple ap proaches are not appropriate.

A numerical technique for the analysis of geotextile reinforced em bankments is outlined. This technique permits consideration of soil‐reinforcement interaction, slip at the soil‐fabric interface, plastic failure within the soil and large deformations. The applicability of the approach has been assessed by examining the observed and predicted performance of a number of reinforced embankments constructed on soft foundations. In this paper, the application of the approach is illustrated by reference to an embankment constructed on a soft peat deposit. Finally, the paper presents a practical design procedure which involves the use of simple design charts. The use of this design procedure is illustrated by means of a worked example using a typical set of design charts.

Resonant column tests were performed on fully saturated and dry hollow cylindrical and fully saturated solid cylindrical specimens of Monterey No. 0 sand of 60% relative density subjected to 96 kPa confining pressure. The hollow specimens were tested by torsional excitation and the solid specimens by longitudinal excitation. It was observed that the threshold shear strain am plitude below which no pore water pressure buildup occurs is approximately $1.5\times {10}^{-3}$ for specimens which were not previously shaken. Between the approximate shear strain amplitudes of $1.5\times {10}^{-3}$ and $8.5\times {10}^{-3}\%$ there was a cyclic strain hardening effect which increased the threshold strain amplitude for specimens which were previously shaken, and caused the pore water pres sure of virgin specimens subjected to cyclic strain to stabilize at a predictable level (liquefaction did not occur). All specimens, whether previously shaken or not, liquefied at the threshold strain of $1.2\times {10}^{-2}\%$ which was previously identified to be the shear strain below which gross particle displacement is frictionally blocked. The effect of previous shaking on damping ratios and shear moduli was also investigated.

The paper presents the results of a full‐scale experiment wherein four embankments were constructed to evaluate compaction equipment per formance, optimum lift thickness, and number of passes required for specified degree of compaction for coal refuse. The coal refuse was mining and prepa ration waste at Moundsville, West Virginia. It was found that suitable equip ment are vibratory smooth drum and sheepsfoot rollers with an applied force not less than 46,000 lb (20,930 kg) and compaction lifts not greater than 1 ft (0.3 m) in thickness.

Data are presented from four projects in which rates of leakage from ponds lined with clay significantly exceeded the rates that would have been predicted on the basis of laboratory permeability tests. The actual hydraulic conductivities of the clay liners were generally found to be 10 to 1,000 times larger than values obtained from laboratory tests on either undisturbed or recompacted samples of the clay liner. The source of difficulty with laboratory permeability tests is the problem of obtaining a representative sample of soil for testing. Neither recompacted samples nor small, undisturbed samples are likely to contain a representative distribution of desiccation cracks, fissures, slickensides, or other hydraulic defects that may be present in the liner. Field permeability tests were performed for three of the four case histories and yielded results that compared well with field performance. Field permeability tests seem to produce much better results than laboratory tests. The four projects had certain features in common: (1) In all cases involving comparatively high rates of leakage, the liners were relatively thin (less than 24 in. or 0.6 m thick); (2) with only one exception, all of the liners were subjected to some desiccation; and (3) in essentially all cases, the construction inspection was not as extensive as it might have been.