Along with the increasing need of construction land, numerous soil reinforcement technologies are proposed in order to improve the soil mechanical properties on one hand and overall site response on the other hand. The presented study is carried out in the context of seismic soil reinforcement and its interaction with a shallow footing which undergoes inertial loading. The system is studied mainly through physical modelling when reduced scale models are constructed in order to simulate clay reinforcement, which is composed of a rigid lower part associated to a flexible upper part. The soft upper part offers shear and moment capacity and the rigid lower part gives bearing capacity. In order to design the reinforcement elements, the response of this combined system to different static and dynamic loads must be understood. This thesis presents results from a primarily experimental study performed in Laboratoire 3S-R (Grenoble). Two reduced (1/10) physical models consisting of a group of four rigid inclusions associated to an upper flexible part are studied in clay. Combined vertical and horizontal static and dynamic loading is applied with a shallow foundation model. A parametric study is done, varying the height of the flexible part of the models in order to define its effect on the settlements of the foundation and lateral performance of the rigid inclusion. A special emphasis was given to the study of the inertial effects of seismic type loading. For this purpose, one of the rigid inclusions was instrumented with 20 levels strain gauges measuring flexural strain, used to calculate the bending moment along the pile. This gives pile deflection (y) by double integration and soil reaction (P) by double derivation. P-y curves are thus obtained. The analysis of the dynamic deflection of the rigid inclusion compared to the movement of the foundation allowed an estimation of the energy dissipated. The results indicate that a large amount of the seismic energy is dissipated within the upper flexible part of the models. Even though the scaling laws are not strictly respected, the main objective of the physical modelling was to perform a qualitative study of the soil reinforcement, studying its behaviour under inertial loading and pointing out important mechanisms, which should be taken into account by the current practice.
