Predicting Tunnel-Induced Settlement in Cohesionless soils in Greenfield Condition

Rapid population growth and increasing congestion worldwide are placing greater demands on existing infrastructure. A solution to this problem is the use of tunnels to construct underground transportation systems. However, tunnelling induces settlement troughs that may result in severe or possibly irreparable damage to nearby infrastructure. For this reason, the effects of tunnelling must be well understood and predicted in the tunnel design stage, to prevent structural damage due to design oversights. Although tunnelling in clay has been extensively studied by previous researchers, the literature lacks reliable settlement estimation techniques for tunnels embedded in granular soil media. In this study, an extensive parametric analysis is conducted to examine the effects of the tunnel diameter, embedment depth, soil density, and shield conicity (tail void gap) on tunnel-induced settlement. Based on the results of the parametric analysis, the authors propose an effective, user-friendly solution that utilizes artificial intelligence software to predict tunnel-induced surface subsidence. The reliability and accuracy of the proposed solution are demonstrated by comparisons with 12 different sets of measured field data.