Safety assessment of shield tunnel underpassing railway bridges considering coupled randomness of stratigraphy and tunneling parameters

During shield tunneling beneath existing railway bridges, the heterogeneity of strata and the uncertainty of tunneling parameters often pose significant risks to railway operation safety. Based on statistical data from geological drilling and shield tunneling parameters in a real engineering project, this study establishes a probabilistic safety assessment framework by integrating unconditional random field theory, Co-Kriging theory, and stochastic function methods. The approach accounts for the spatial distribution constraints of geological exploration data and the statistical characteristics of tunneling parameters to simulate the coupled stochastic effects of stratigraphic physical-mechanical properties and tunneling process parameters. Numerical results reveal that simulations considering both stratigraphic information and tunneling parameter randomness exhibit significantly greater variability compared to those considering only stratigraphic or tunneling parameter randomness, underscoring the importance of their coupled stochastic effects. Furthermore, the proposed method evaluates the probability of exceeding deformation limits under various control values for railway bridges. For the case study, no speed restrictions are required during left-line tunneling. However, for right-line tunneling, a speed limit of 80 km/h is necessary when the shield head reaches beneath the railway bridge, reducing to 45 km/h as the shield moves away from the bridge. The most critical phase occurs when the shield tail exits from beneath the bridge during right-line tunneling.