Understanding surface settlement in mechanized tunneling through integrated analytical and predictive approaches

Surface settlement is one of the major challenges in mechanized tunneling, as it can significantly affect ground stability and existing infrastructures. Despite extensive research on surface settlement in mechanized tunneling, the combined investigation of parameter sensitivity using GRA, the numerical assessment of shield conicity, and the comparative evaluation of mathematical prediction models based on real large project data has received limited attention in previous studies. This study investigates the parameters influencing surface settlement using Grey Relational Analysis (GRA) method, evaluates the effect of the shield’s conical configuration through numerical modeling, and estimates the long-term surface settlements using mathematical models. A comprehensive quantitative dataset obtained from the Tabriz Line 2 Metro project was employed to perform a data-driven analysis and enhance the reliability of the results. The GRA results indicate that tunnel depth is the most influential parameter affecting surface settlement. Based on these findings, key parameters were selected for numerical modeling, and the impact of the conical shield geometry was analyzed under three different geotechnical stratifications. The numerical simulations using PLAXIS show that the shield passage can induce settlements exceeding 40% in soils with varying structural characteristics. In addition, long-term surface settlements were predicted using Semi-logarithmic, Hyperbolic, and Hybrid mathematical models. The comparative analysis demonstrates that the Semi-logarithmic approach provides more accurate predictions compared to the other methods. The results highlight that integrating data-driven analytical techniques with numerical modeling can significantly improve the understanding and prediction of surface settlement in mechanized tunneling projects. The novelty of this paper emphasis on the utilization of mathematical correlations in the log-term settlement magnitudes which is based on a large field data and can reduce both project costs, controlling time, and possible design or construction risks.