Study on Smoke Spillover Behavior in Subway Train Fires at Different Fire Source Locations and Its Effects on Tunnel Ceiling Temperature Distributions

This study employs Fire Dynamics Simulator (FDS) to systematically investigate smoke spillover behavior and thermal characteristics in subway train-tunnel coupled spaces under varying fire source locations and heat release rates. We focus on the maximum temperature rise along the curved tunnel wall, longitudinal and transverse ceiling temperature distributions, and the evolution of smoke layers. Results indicate that the maximum temperature on the curved wall occurs at approximately 40° below the tunnel centerline. Fire source location significantly modulates smoke spillover dynamics; fires directly facing doorways cause concentrated discharge and higher peak wall temperatures, whereas end-wall constraints promote longitudinal propagation within the carriage, resulting in relatively lower tunnel temperatures. Notably, end-carriage fires exhibit a secondary longitudinal temperature peak due to smoke accumulation and recirculation near the terminal region, contrasting with the approximately symmetric distribution observed in middle-carriage fires. Based on mass conservation principles, a transverse temperature decay model is derived for the tunnel wall at varying angular positions, with corresponding Stanton numbers determined through regression analysis. Furthermore, we reveal that smoke discharge is dominated by longitudinal inertia over transverse inertia, leading to highly non-uniform spillover patterns. This study clarifies the coupled evolution mechanisms of smoke spillover and temperature fields in subway train fires, providing a robust theoretical basis for temperature detection, smoke exhaust system design, and emergency evacuation safety assessment.