Parametric modeling and safety simulation of pit excavation affecting adjacent tunnels based on BIM-FEM framework

Abstract

This research puts forward an intelligent analysis framework to tackle key issues in the collaborative analysis of Building Information Modeling (BIM) and the Finite Element Method (FEM) within urban foundation pit projects. These issues include model conversion distortion, poor mesh convergence, and low efficiency in multi – case iterations. By integrating BIM parametric modeling with FEM automated analysis, this framework enables the deformation analysis and safety assessment of existing tunnels during foundation pit excavation. Firstly, a 3D geology – support – tunnel parametric collaborative modeling workflow on the BIM platform is utilized to enhance collaborative design efficiency and facilitate dynamic updates. Secondly, through the secondary development of TCL (Tool Command Language) and Python scripts for the meshing software Hypermesh and the numerical simulation software ABAQUS, a hexahedral mesh optimization algorithm and an automated pre-processing process are established. This effectively resolves problems of geometric aberration and mesh convergence in traditional BIM-FEM data conversion. Finally, an interactive platform for multi-case analysis is employed to analyze the spatio-temporal evolution of neighboring tunnel deformation during excavation. Validation using typical soft-soil pit projects reveals that the simulation error for tunnel deformation is maintained within the range of 1.36% to 4.7%, while the errors for surface settlement and ground wall displacement remain below 6.31%. The displacement characteristics of the ground wall in the soft – soil layer are captured with an average accuracy of 97.45%. The study shows that the horizontal and vertical displacements of the tunnel decay exponentially with the distance from and depth of the foundation pit. It identifies the deformation – sensitive areas of existing tunnels and sets up critical position control parameters for displacement thresholds. This framework overcomes the efficiency bottlenecks of traditional manual modeling, reducing the time for a single design change from 8 to 12 h to less than 30 min. Thus, it offers a high – precision and efficient solution for safety assessments in sensitive surrounding environments through digital twin technology.