This study investigates the application of Ordinary Stone Columns (OSCs) and Geogrid-Encased Stone Columns (GESCs) in enhancing the properties of soft clay soils through numerical analysis using PLAXIS 3D (version 2024). The study contrasts numerical findings with two well-researched field case studies: one in Korea and one in Iraq. The analyses were calibrated using the Mohr-Coulomb and Hardening Soil models, and settlement responses were assessed for different reinforcement scenarios, including untreated soil, OSCs, and GESCs. The results show a strong match between PLAXIS 3D simulations and field measurements, confirming the method's reliability. In the floating case (in Iraq), OSCs increased load-bearing capacity by about 21%, while GESCs improved it by around 30% compared to untreated soft clay. For the end-bearing case (in Korea), even greater enhancements were recorded, with OSCs increasing the bearing capacity by nearly doubling it and GESCs by almost 2.5 times compared to untreated soil. Geogrid encasement is presented as significantly improving settlement control and bearing capacity, with PLAXIS 3D proving to be an important design aid in geoground improvement systems.
This paper offers a comprehensive review of column-based ground improvement techniques, focusing on their fundamental mechanisms, design principles, construction methods, and field applications. It highlights stone columns and deep soil mixing (DSM) as the most widely used and effective solutions for enhancing the performance of weak and compressible soils. The core principles, including stress redistribution, increased shear strength, and accelerated consolidation, are discussed in detail. The review synthesizes key design parameters such as column geometry, area replacement ratio, and the role of geosynthetic reinforcement and load transfer platforms. It also examines the practical application of these methods through various case studies on embankments, tank foundations, and excavation supports. A dedicated section explores the pivotal role of numerical modeling, especially the finite element method (FEM), and emerging AI-driven approaches like Physics-Informed Neural Networks (PINNs) and surrogate modeling, which are shown to improve predictive accuracy and optimize the design process. Furthermore, the paper addresses critical challenges and limitations, including material variability, installation uncertainties, environmental impacts, and the need for enhanced quality control and long-term monitoring. It concludes by outlining future trends and innovations, such as the adoption of sustainable materials and the integration of machine learning for predictive design and real-time monitoring. This synthesis provides a structured overview of current best practices and offers valuable insights into the future direction of this vital area of geotechnical engineering.