Limitation of convergence-confinement method on three-dimensional tunnelling effect

The convergence-confinement method (CCM) is a simplified widely utilised tool for assessing the interplay between the rock mass behaviour and the support effect, so it is quite helpful for tunnel support design purposes. However, the direct application of this technique has shown some limitations, many of which are directly related to the three-dimensionality issue. Indeed, the CCM tries to solve the three-dimensional (3D) problem of tunnel advance deformation and support response, by means of a series of two-dimensional (2D) plane strain analyses. So, regardless purely elastic cases, certain deviation is observed when comparing CCM and 3D numerical modelling results. The reasons behind this deviation have been studied from different points of views, but they seem to be still not well understood. With the aim of advancing towards a better knowledge of this issue, this paper discusses the limitation of CCM to correctly reflect the 3D tunnelling effect by comparing CCM and 3D numerical deformation, support pressure and liner load results in a typical tunnel case for various geological conditions. The reasons for CCM results in different rock deformation and support pressure in comparing to the 3D numerical modelling are explained. Some guidelines are eventually given recommending when the use of CCM can be acceptable according to the rock mass strength and tunnel depth, and when a more rigorous 3D approach is convenient.