Probabilistic analysis of the thermal insulation length demand of tunnels in cold regions via the FEM-CFD coupled method

In winter, cold air flowing into tunnels undergoes heat exchange with the surrounding rock and lining, rendering frost protection a critical concern. In this paper, a practical procedure for integrating the FEM–CFD coupled method into probabilistic analysis is presented for determining the thermal insulation length demand of tunnels in cold regions, in which uncertainties in the air temperature and wind speed are considered. The results indicate that with ventilation implemented in the coldest month, tunnel temperatures gradually rise along the longitudinal direction, exhibiting a distinct temperature variation near the entrance. Based on the numerical simulation results obtained with 100 air temperature and wind speed samples, generated using the Monte Carlo method, the thermal insulation length demand of the tunnel ranges from 1000 to 1500 m, showing a distribution that follows the characteristics of a normal distribution. Additionally, a corresponding surrogate model is developed, and a thermal insulation failure probability model is proposed to analyze the sensitivity of meteorological parameters, revealing that the air temperature is the primary factor influencing the thermal insulation length demand. The findings of this research provide novel insights for the engineering community in designing resilient and sustainable tunnels in cold regions, mitigating the potential risks associated with freezing events and ensuring safe and efficient transportation infrastructure.