A Comparative Study of Greenfield Tunnelling in Sands: FEM, DEM, and Centrifuge Modelling

Abstract

The effect of tunnel construction on ground displacements is an important problem for tunnelling engineers. Numerical methods, including continuum and discrete element methods, have been used to evaluate tunnelling induced ground displacements. In this paper, the ability of numerical methods to replicate the response to tunnelling of a real soil is evaluated by comparing results from numerical analyses with experimental data obtained from geotechnical centrifuge tests. The centrifuge tests include two types of tunnel boundary condition: pressure-controlled (water extracted from model tunnel within a flexible membrane) or displacement controlled (rigid boundary model tunnel undergoing an eccentric contraction). Centrifuge measurements are compared against discrete element method (DEM) and finite element method (FEM) analyses which replicate the conditions of the experiments: pressure controlled boundary for FEM and DEM; displacement controlled boundary for FEM only. The effects of tunnel boundary condition on the soil displacement mechanisms are illustrated and the performance of the numerical analyses to replicate salient features of ground response are discussed.