Numerical investigation and statistical optimization of geotextile-reinforced embankments over soft coastal deposits: Integrating FEM and RSM analysis

Embankments constructed over highly compressible soft coastal clays frequently encounter severe stability challenges due to low shear strength and excessive settlement, particularly in high-rainfall regions such as Teknaf, Bangladesh. This study presents an integrated computational framework that couples Finite Element Method (FEM) simulations in Plaxis2D with Response Surface Methodology (RSM) to conduct a systematic parametric investigation and statistical optimization of geotextile-reinforced embankments. Seventy-two numerical simulations were executed by varying embankment height (2-5 m), slope angle (14°-45°), and the number of geotextile reinforcement layers (0-5), based on realistic soft coastal clay parameters. The numerical models were validated against the Limit Equilibrium Method, demonstrating strong agreement with discrepancies ranging from 1.0% to 6.8%. The results indicate that geotextile inclusion substantially enhances the safety factor, particularly for higher and steeper configurations, with marginal performance gains decreasing beyond three reinforcement layers. Furthermore, the developed second-order RSM model demonstrated good predictive capability (R 2 = 0.9332) and captured important parameter interactions and geometric effects within the studied design space. Within the Teknaf coastal clay conditions considered, the integrated FEM-RSM framework provides predictive relationships and design-oriented insights that may support more cost-effective embankment configurations over soft coastal deposits.