No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

Considerable efforts have been made in recent years to develop a better understanding of the mechanical behaviour of unsaturated soils in terms of elastoplastic critical state constitutive models. These models are defined in terms of four independent stress state variables : mean net stress, deviator stress, suction and specific volume. An important feature of the models is the suggestion of the existence of a yield surface in mean net stress : deviator stress : suction space. Suction-controlled triaxial tests were performed to investigate the shape of the yield surface for unsaturated compacted speswhite kaolin with a particular stress history. The tests were conducted in a Bishop-Wesley triaxial cell with suction applied by the axis translation technique. The soil samples were instrumented with local strain gauges for measuring the sample volume change. Ten samples were tested, and in each test the location of the yield surface was fixed by isotropically consolidating the sample to a mean net stress of 400 kl'a and a suction of 100 kPa. This procedure effectively erased the one-dimensional stress history produced by the compaction process. After unloading to stress states inside the yield surface, samples were re-loaded either isotropically or by drained triaxial shearing. Re-loading stages were conducted at three different values of suction. Yield points in the re-loading stages were most easily identified from plots of specific volume against the logarithm of mean net stress (even for shear tests), whereas it was often difficult to identify a clear yield point from a plot of deviator stress against shear strain. The yield points identified from the re-reloading stages were used to define the shape of the yield surface in a stress space with axes of deviator stress, mean net stress and suction. Constant suction cross-sections of the yield surface were approximately elliptical in shape, with one axis of the ellipse coinciding with the mean net stress axis. As expected, the size of the elliptical constant suction yield curves increased with increasing suction Soil elastic indices obtained from swell-back and suction-change stages showed non-conservative behaviour, indicating that the behaviour of the soil was not truly elastic even for stress states inside the yield surface. Plastic strain increment vectors were plotted for the post-yield behaviour, and these were consistent with an associated flow rule.

The landscape of Northern England has been mostly formed by glacial activities during the Quaternary period, and glacial till materials have been deposited over the northern counties of England during these glacial activities. Townships, industrial developments and infrastructure works exist or are planned in these areas. The variable and often complex successions in which glacial tills occur have frequently led to problems on civil and mining engin eering projects. Glacial tills are engineering soils which have been defined as a poorly sorted mixture of clay, silt, sand, gravel, cobble and boulder sized material deposited directly from glacier ice. The glacial tills of the counties in Northern England are the subject of many studies which are carried out in order to determine the properties of the overlying glacial deposits. Ground investigations have been carried out for opencast coal projects. A large number of samples were obtained and extensive laboratory testing has been carried out. Using the results of these investigations and tests, a geotechnical database is being developed that should provide a useful resource for civil and mining engineers in the northern counties region. Its purpose is the extensive analysis of the parameters that are used to define the geotechnical properties of Northern England glacial tills. This should give a better understanding of the engineering behaviour of glacial tills and parameter selection for engineering design. In addition to statistical analysis, Neural Networks, a model of Artificial Intelligence, are used to find correlations between the different parameters and to develop new methods of modelling and predicting geotechnical design parameters. Neural technology is an emerging field of artificial intelligence that has attracted the interest of many scientists and engineers. They are information-processing systems that can mimic the biological system of the brain and can be trained to complete and classify input patterns, or to complete a function of their input. In this project the data available from the database are used to train Neural Networks to classify glacial tills according to their geotechnical properties and investigate their potential in predicting geotechnical design parameters.

No description available.