Random Field-Based Tunneling Information Modeling Framework for Probabilistic Safety Assessment of Shield Tunnels

Abstract

Probabilistic analysis based on random field (RF) has been widely adopted in the safety assessment of shield tunnels. However, its practical applicability has been limited by the intricacy involved with integrating geotechnical data and tunneling information. This paper addresses the following research question: How can the RF-based probabilistic safety assessment be carried out efficiently? In addressing this research question, we suggested an RF-based tunneling information modeling (TIM) framework to realize the probabilistic safety assessment of shield tunnels. In the proposed framework, the modeling of tunnel structure and geological conditions is initially introduced. The numerical safety assessment model is then created via an automated procedure using the RF-based TIM. A case study is conducted to verify the suggested framework, and results demonstrate that the framework can offer an automated design-to-analysis solution to improving the safety assessment of shield tunnels by comprehensively considering the uncertainties of geological conditions.