Objective  The primary technical objective of the present work is to combine real data in terms of a subway tunnel structural health (including its concrete structure and the surrounding soil and rock support and burial) at strategic locations along a segment of the tunnel obtained by fiber optic sensors of strain and displacement, and simulations based on finite element continuous mechanics, and to assess the sensor effectiveness as well as its deployment locations. Iteration of this process leads to an optimal sensor system for the effective and efficient monitoring of the subway tunnel under both static and dynamic operating conditions. A structural health monitoring project of geotechnical application is carried out based on the real-life subway segment of the Suzhou Metro Line 4 branch from Hongzhuang Station to Lishu Station, using finite element simulation analysis of the static stress/strain field and dynamic field of tunnel structures under normal operating conditions and deploying optical fiber sensor network.

  Methods  Simulation and modeling are carried out using standard finite element analysis of continuous mechanics; Experiments are conducted using a system of fiber optic sensors, including strain gauges, displacement sensors. The data output of the sensors over an extended period of time is used to analyze the structural health of the tunnel, and the simulation/modelling is compared with the sensor data, which also serves to guide the placement of the sensors. According to the simulation results and on-site investigation results, a fiber optic sensing network which can monitor real-time information such as horizontal displacement, cracks, settlement, structural stress, surface strain, and vibration of tunnel structures is designed and deployed successfully. The real-time health status monitoring of tunnel structures is achieved and the reliability of the finite element model is verified, which resulted in a more accurate finite element simulation model. Finally, two-dimensional finite element models are established for three main sections of the subway segment, and the distribution of concrete structural damage under the influence of metro train vibration loads is studied under two conditions: with and without damage to the tunnel structure.

  Results and Discussions  The main results of this study include: 1) Statics analysis points to the fact that the maximum stress of the tunnel is located at the lower part of the tunnel, with a symmetric distribution, and the maximum strain is located at the top of the tunnel. 2) For the dynamics analysis, the resonant frequency and vibration modes were first obtained, based on which the acceleration and velocity distribution were obtained. 3) Using the simulation obtained static and dynamic distribution and actual field observations, fiber Bragg grating displacement sensors, strain sensors, crack sensors, static pressure sensors and distributed fiber optic vibration sensors were place, for long-term, in-situ monitoring of the tunnel structural health. 4) Based on the measurement output of the system of fiber optic sensors, the finite element model is updated and optimized, to reflect the real situation of the tunnel.

  Conclusions  As a result of the work reported, better insurance of the safe operation of the subway segment has been achieved, and the model and methods developed here can be readily applied to other subway segments, and more generally other geotechnical structures involving underground tunnels and tracked vehicular traffic. The present study has reported the successful use of a combination of finite element static and dynamic modeling with fiber optic sensor network output, for better insurance of the safe operation of underground subway tunnel. The model and methods developed and demonstrated here can be readily applied to other subway segments, and more generally other geotechnical structures involving underground tunnels carrying tracked vehicular traffic.