An adequate representation of model uncertainties is important for reliability analyses based on geotechnical analysis models of unknown credibility. A procedure is outlined for calibration of the probability distribution function for the model uncertainty associated with such a geotechnical model. The calibration is based on available field test measurements of quantities, which can be predicted by means of the actual model. The procedure is generally applicable, provided a sufficient amount of field test measurements is available. The procedure is outlined with specific reference to a prediction model for axial pile capacity and a set of pile capacity measurements from field tests on full‐scale piles. An example is presented to demonstrate the applicability of the procedure, based on the American Petroleum Institutes's method for prediction of shear capacities of piles. A first‐order reliability analysis of an example pile is carried out and demonstrates the significant importance of a best‐possible representation of model uncertainty in geotechnical reliability analyses.

This paper concisely presents a new method for the estimation of settlements of shallow foundations in cohesionless soils. The method includes a simplified model for the distribution of applied stresses with depth, incorporates certain experimental results concerning the stress‐strain relation in cohesionless soils, and proposes a formula for selection of the effective depth below which the deformations can be considered negligible. The main conclusions derived for the experimental investigation on various types of cohesionless soils are used in the analytical model for the estimation of settlement. The final formulas for the calculation of settlement, although they have the form of the elastic‐type solutions, are, in effect, nonlinear functions of the foundation width and of the applied loading. The influence of certain basic factors on the settlements of cohesionless soils, such as, the effect of stress history and the width of the foundation, is examined. It is concluded that the trends from known statistical correlations are predicted satisfactorily by the proposed method. Back‐calculations of settlement using data from published cases, also showed generally satisfactory approximation.

A probabilistic model for stability analysis of a deep‐water, gravity‐ based platform foundation on clay is defined, based on theory for wave loading in conjunction with conventional shear surface considerations. An advanced total stress analysis method is applied for representation of the undrained shear strength of the clay, and simple modifications to account for effects of strength degradation caused by cyclic loading are included. The probabilistic model accounts for uncer tainties in loading as well as in soil properties, and model uncertainties also are considered. The practical application of the model is illustrated by presentation of calculations for a typical foundation, and the foundation of a concrete monotower at a Norwegian Trench location is considered as an example. This foundation consists of a base slab, resting on the sea floor, in combination with vertical skirts penetrated into the clay at the site. Probability calculations are performed by a first‐order reliability method. A sensitivity study is included.

An extensive reevaluation of the seismically induced 1971 Lower San Fernando dam slide is performed using steady‐state strength concepts. This investigation includes both field and laboratory testing and is carried out as a cooperative effort by four organizations. The undrained steady‐state strength $\left(S_{\mathrm{us}}\right)$ of the hydraulic fill is evaluated based on laboratory testing of undisturbed samples obtained (1) From boreholes; and (2) by hand trimming in a large‐diameter exploratory shaft. The resulting $S_{\mathrm{us}}$ values are corrected for the effects of void‐ratio changes due to sampling and testing, as well as for void‐ratio changes due to the 1971 earthquake shaking. The strength of the hydraulic fill is also backfigured from the extent of the slide movements and compared with the measured (corrected) strengths; the results indicate that the strength mobilized during the failure is about equal to the average minus about one half to the average minus one full standard deviation values of the measured (corrected) laboratory strength estimates.

A constitutive model using multiple yield surfaces is developed for analyzing foundation problems on anisotropic soils. The model parameters are obtained from simple triaxial test data and a close model prediction compared to the experimental data for two different soils are obtained. A nonlinear finite element analysis of two‐dimensional plane strain footings and three‐dimensional footings is done and the influence of anisotropy on the bearing capacity and settlement are studied. For the range of anisotropy encountered in the field, the bearing capacity and settlements for anisotropic soils can differ by over 100% compared to the isotropic case. Anisotropy has the largest influence in strip footings and the lesser in square footings. The constitutive model developed is simple in form and easy to use and yet very general to accommodate the complex anisotropy of the soils.

A series of trench backfill tests are performed at Duke Power Company's Marshall Steam Station to evaluate the performance of a flowable fly‐ash backfill versus that of a typical compacted soil backfill. The test site consists of a series of parallel trenches traversing a service road used for hauling fly ash to an on‐site storage area. Instrumented pipe assemblies are installed in trenches that are backfilled using both compacted soil and a previously designed fly‐ash slurry. Pipes of various sizes and materials are instrumented with strain gages to monitor longitudinal and circumferential strains under static and dynamic loading conditions. Instrumented load links are also installed at selected pipe joints to evaluate the relative resistance to joint opening for both types of backfill. Static and dynamic loads are applied by heavy‐vehicle traffic during normal hauling operations. In this paper, is part one of two comparative pipe strain data are presented for static loading conditions. The effects of additional compaction due to normal traffic and subsequent regrading are evaluated for the soil backfill, and strain data are presented for the fly‐ash slurry backfill. The project was designed to demonstrate the feasibility and safety using a flowable fly‐ash backfill in an environment subjected to continual vehicular loading.

This is the second paper in a two‐part series describing trench backfill tests designed to evaluate the performance of a flowable fly‐ash backfill (ashcrete) versus that of compacted soil backfills. The site for the tests is Duke Power's Marshall Steam Station. For these tests, a series of parallel trenches is installed in the outbound lane of an on‐site service road. The major use of this road is for the hauling of fly ash to an on‐site storage area. Instrumented pipes are installed in the trenches, which are subsequently backfilled using both compacted soil and ashcrete. Surface strains for the buried pipes and joint loads are monitored for static and dynamic loading conditions. The monitored dynamic strains and joint loads resulted from normal heavy‐vehicle traffic at the site. Comparative surface strain and joint‐load data are presented at two stages of compaction for the soil backfill and the ashcrete backfill. Additional strain data for the ashcrete backfill are included to demonstrate the continued performance of this backfill after additional curing. The project was designed to demonstrate the feasibility, advantages, and safety of using this flowable fly‐ash backfill in a typical continuous high‐load application.

This paper documents a low‐cost, qualitative evaluation scheme using a fuzzy set analysis for mapping slope failure potential (SFP). Four categories of factors affecting stability of natural slopes, i.e., geology, topography, meteorology, and environment, are considered. Each category involves two to five factors, and a total of 13 factors is used in the proposed evaluation scheme. A set of evaluation criteria was established for each adopted factor. In the proposed evaluation scheme, slope failure potential is assessed and recorded in linguistic terms based on the established criteria. The linguistic data or information obtained from the assessments is represented and processed using fuzzy sets. In this study, the analysis or computation involving fuzzy sets was performed using the Monte Carlo simulation technique. An SFP index to measure the slope failure potential is defined. Based on the computed SFP values for a large number of subareas or cells selected for an area studied, an SFP map can be developed. An example application is presented to demonstrate the method. The SFP map developed in the case study seems to be able to correctly predict the slope failure potential of the area studied.