Effects of the excavation pattern on stress reduction factor in convergence confinement method in soft ground- case study Karaj metro line 2

Summary Tunnel construction is in demand in many large cities in the world that facing a problem with traffic. Among 2D methods to tunnel analysis, the convergence–confinement method (CCM) is the most accurate method that analyzes the tunnel-ground-support interaction. Previous researchers’ studies have been done mostly to evaluate the affected parameters on CCM. But few of them are investigated how effected the shallow tunnels (anisotropic stresses) with any cross section shapes to the confinement loss values in all around perimeters of tunnel section.
Introduction The aim of this study is to evaluate the effect of excavation pattern on stress reduction factor in shallow tunnels with horseshoe cross section shape in soft ground
Methodology and Approaches For this purpose, at first, by considering the line two Karaj subway cross sections, using back analysis, calibration of Mohr-Coulomb constitutive model parameters was determined. Then, by considering common type of excavation patterns (full face and multi face), the three different curves, ground reaction curve, longitudinal displacement profile and support characteristic curve, which is required to obtain the stress reduction factor. This is done using a series of numerical analysis by FLAC 2D and FLAC 3D code. Back analysis is done to obtain soil parameters of Karaj subway line two. The stress reduction factor values are investigated in each excavation patterns in different points around the tunnel. Finally, the vertical and horizontal movement induced mentioned effects are evaluated in each tunnel stages.
Results and Conclusions The results of this study show the good agreement of parameters obtain from back analysis to instrumentation data. Tunnel excavation pattern has significant effects on determining stress reduction factor in convergence confinement method for multi stage tunneling excavation. The results of analysis are shown a good capacity of model in reproduction of better and more precise tunnel behavior in shallow tunnels.