Physical modeling of stone column installation in fine sand and its effect on lateral to vertical stress ratio Olivier Hurley, Mathieu Nuth & Mourad Karray Department of civil engineering – Université de Sherbrooke, Sherbrooke, Québec, Canada Frédéric Massé & Claire Mollereau Geopac inc., Boucherville, Québec, Canada Mounir Bouassida Université de Tunis El Manar – École Nationale d'ingénieurs de Tunis (ENIT), Tunis, Tunisia Mohamed Chekired Institut de recherche d'Hydro Québec (IREQ), Varennes, Québec, Canada ABSTRACT An experimental research program has been undertaken to investigate the lateral response of stone column installation. A new apparatus was developed to apply a vertical stress on a sand sample inside a 600 X 364 X 12 mm PVC cell that allows for the installation of a stone at its center. Lateral pressures were measured using a Tekscan flexible sensor and lateral to vertical pressure ratios were measured before and after the stone installation to evaluate the inherent improvement. The experimental procedure also included density tests using the dynamic penetration method. The results obtained showed increased sand density and increased lateral to vertical pressure ratio for all three of the stone columns tested. RÉSUMÉ Un montage expérimental est développé dans le but d'étudier la variation de la pression latérale suite à l'insertion de pierre dans un sable. Un appareil est construit à cet effet permettant l'insertion de pierre tout en maintenant une surcharge verticale sur un sable à l'intérieur d'une cellule ne PVC de dimensions 600 X 364 X 12 mm. La pression latérale est mesurée à l'aide d’un capteur de pression Tekscan ultramince et flexible. La procédure d’essais comprend également des essais de densité basés sur la méthode de la pénétration dynamique. Les résultats obtenus montrent une augmentation de la densité du sable et du ratio des contraintes latérale et verticale pour les trois colonnes ballastées testées. 1 INTRODUCTION The vibro-stone column method consists of inserting a stones using vibration and repeated downward and upward movement of the probe into soft or weak soil. Although vibration is often considered ineffective in cohesive soils due to undrained conditions during installation (Castro, 2010), its effect is considerable in granular soils. The improvement of weak soils due to stone column installation can be measured in terms of bearing capacity and settlement reduction, but also in reduction of liquefaction potential where applicable (Barksdale, & Bachus 1983 - Mitchell, 2008). In order to understand the actual improvement of weak soil using vibro-stone columns and the installation effect mechanisms have on supporting or encasing soil must be properly understood. The best way to evaluate the increased strength of the soil is by measuring the lateral to vertical pressure ratio known as the coefficient of earth pressure at rest K0 when no external forces are being applied. A well-known analogue for the stone column is the use of a unit cell model where analysis of a single stone column can represent an infinite mesh of stone columns. This concept is applicable for uniformly distributed loads and where the symmetry of the deformations are null at the cell extremity and uniform at the top according to Balaam & Booker, (1981). A custom designed apparatus shown in figure 1 was developed at the Université de Sherbrooke which allows for the insertion of stones using a similar method to the Franki pile method (Barksdale, & Bachus. 1983) with an applied load of up to 300 kPa. This method omits the effect of vibration during vibro-stone column installation and focuses on the lateral pressures and density variations due to the installation process. The apparatus is compatible with the unit cell concept. The experimental program included the installation of three separate stone columns where the total lateral and vertical pressures were measured and compared to determine the K ratios. A dynamic penetration test (DPT) was developed in order to determine the relative density of the soil at 3 different distances from the axis of the column. 2 LABORATORY EXPERIMENT 2.1 Soil characterization The soil to be improved was fine sand with silt and the stone used for the column was uniform fine gravel. Laboratory tests were performed to determine mechanical