Foreword

Embankment construction on soft soils is accompanied by significant settlement and risk of bearing capacity failure. Ground improvement offers significant economic advantages, ranging from rigid inclusions, such as piles, to more flexible options, such as stone columns or sand compaction piles. When combining the latter optimally with a base geosynthetic layer under the embankment, load transfer will be improved, Settlements reduced, failure prevented and the consolidation process will be accelerated. This dissertation explores how the various elements interact, as embankment loads and Settlements increase towards external and internal failure, through bearing capacity, slope stability or stress concentration in the stone columns, respectively.

Physical and numerical modelling of the entire soil-embankment System, as represented by a given prototype model, was conducted, beginning with installation of stone columns through to the placing of an embankment and dissipation of the excess pore pressures in the soft soil. There was significant innovation during this study, in particular in the physical modelling in a geotechnical centrifuge, and in the post-test laboratory and imaging investigations. Dr Weber led the construction and implementation process for a device to construct the stone columns in-flight, to reproduce the stress history of the soil. The radial compression in the soft soil and a smear zone in the annulus around the column were replicated through repeated compaction of the granulär material in the excavated hole, resulting in increased lateral earth pressure in the surrounding soil. Parametric studies were then carried out in the centrifuge strong boxes and the füll drum centrifuge, with untreated as well as improved instrumented soft soil sections, loaded either to working conditions or to failure.

Smear zones around the stone columns were investigated at particle scale through a combination of mercury porosimetry and environmental scanning electron microscopy, to determine the change in porosity and density of the clay immediately adjacent to the column. Various zones were identified around the stone column as a result of installation. Of greatest significance was a residual shear zone was caused by penetration of the construction tool, which could reduce both the axial bearing capacity of the sand columns and the radial permeability essential for promoting accelerated pore pressure dissipation.

Numerical modelling using finite elements replicated the installation processes within a Single unit cell analysis to simulate the placement of the column and subsequent compaction, as well as representation of the pore pressure, settlement, System load transfer, stress concentration in the stone columns and the relative Settlements within the sand column and surrounding clay. Parametric studies were carried out to determine the influence of the reinforcement degree. Subsequently, a twodimensional plane strain approximation of the three-dimensional stone column System was compared with the centrifuge model test carried out in the Container.

Dr Thomas Weber has completed a particularly challenging research project with notable success, resulting in delivery of several innovative Solutions to this problem of engineering significance, both in Switzerland and further afield. Insights obtained from this thesis make a valuable contribution to understanding the interacting mechanisms and will undoubtedly be applied in future in practice. In the meantime, Dr. Weber's developments have been adopted for further doctoral studies into ground improvement methods for problematic soils.

Prof. Dr. Sarah Springman OBE