Publication: Evaluation of soil-twin tunnel interactions in kenny hill formation using numerical approaches
Date
2020-09-01
Authors
Govindasamy, Darvintharen
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Tunnel development provides better solutions in contrast to surface development, but settlement induced from tunnelling activity potentially triggers the ground movement problem to the surrounding. Finite Element Method (FEM) is one of the widely used numerical methods. Therefore, adopted in this study to predict the settlement that produced above the shallow twin tunnels of Sungai Buloh to Kajang (SBK) Line in Kenny Hill formation. This study highlights the simplified 2D FEM approaches to predict tunnel settlement. Initially, the 3D ground models are developed for subsurface characterisation. Five tunnel cross-sections that selected based on the availability of ground monitoring data is selected for the numerical analysis from the 3D SPT-N ground model. Hardening Soil (HS) model primarily used as the soil constitutive model. The stiffness parameter of the HS model is established from various empirical correlations. From the effectiveness evaluation of empirical correlation, the 1.74 N (N referring to SPT-N) relationship is chosen to adopt throughout the analysis. A comparison between the HS model and Mohr-Coulomb (MC) model is conducted in the study where the parameter of the MC model is extracted from the HS model. 2D FEM simulation is carried out in the Plaxis 2D using the contraction method, stress reduction method and modified grout pressure method. Back analysis approach is performed to match the predicted settlement trough with the ground monitoring data of twin tunnels. All the methods are able to provide good agreement with the ground monitoring data including an average different less than 0.248 mm. The contraction ratio, unloading factor and face pressure achieved in the study area is ranging from 0.16 % to 0.92 %, 0.16 to 0.61 and 205 kN/m2 to 348 kN/m2. The ratio of calculated face pressure to measured is 1.13 to 1.89. The comparison between the MC model and HS model proves that the HS model able to perform well in terms of settlement prediction and behaviour compared to the MC model. Relatively good correlation between the unloading factor and contraction ratio with R2 of 0.944 achieved in this study. Furthermore, an inter-correlation between the three methods can be used as a guidance to predict the unloading factor values or the face pressure with a specified percentage of contraction ratio or even volume loss and vice versa. At last, a simple 3D FEM is conducted in one selected tunnel section and the result is compared with the 2D FEM output. The 2D FEM is very useful in solving tunnel induced settlement problems, in particular those induced by twin tunnel interaction.