The influence of fabric on mechanical behaviour is explicitly considered in some of the current constitutive models for unsaturated soils. These are based on assumptions regarding the interaction between different levels of structure, which still require experimental validation. A study has been carried out to investigate the evolution of fabric in a compacted natural clay during wetting and loading, and the results are presented in this paper. Samples of London Clay were statically compacted to the same initial conditions, dry of optimum moisture content in a Proctor plot, and subsequently taken along complex stress paths involving wetting, loading, or a combination of both. Mercury intrusion porosimetry (MIP) and environmental scanning electron microscopy (ESEM) microstructure investigations were carried out to observe and quantify the change in fabric associated with each path. The soil fabric was observed to change from an aggregate to a matrix type structure along all wetting paths. This transition was found to take place only when the suction was reduced to a value close to zero kPa. Results also showed that fabric changes during yielding were stress path-dependent. It was not possible to find a correspondence between the volume of free porosity and the volume of intra-aggregate pores, as suggested by some authors. Finally, the volume of pore water was found to agree closely with the volume of intra-aggregate pores, providing support to the assumption that in an unsaturated aggregate microstructure the clay aggregations are saturated.

No abstract

No abstract

The understanding of partly saturated soil behavior can be enhanced with the help of additional specialist testing in the laboratory. From the various techniques available to do so at present, the osmotic technique is probably the simplest one. However, there are concerns with the response of this system over time, and in particular its ability to sustain suctions for long periods of time. In this study, the response of a synthetic semi-permeable membrane in conjunction with a solution of high molecular weight has been investigated. The behavior of the osmotic system has been studied both in isolation and in full scale tests, by incorporating it into a purposely built oedometer. Not only was it possible to establish a calibration curve of osmotic potential versus concentration of solution but, additionally, it was possible to perform tests which extended for up to five months. The results suggest that with the correct choice of semi-permeable membrane and solution, it should be possible to use the osmotic technique in a satisfactory manner to perform any desired test on partly saturated soils, regardless of its duration.

No abstract