Subway trains cannot stop immediately to extinguish fires and evacuate passengers if a fire accident occurs. The piston wind generated by train movement coupled with longitudinal ventilation, makes the process of high-temperature smoke spreading upstream and downstream of the fire source more complex and unpredictable. The evacuation process of personnel is affected by high-temperature smoke in tunnels, it is worthwhile to investigate personnel evacuation in interval tunnels under longitudinal ventilation during a train fire. This paper uses a three-dimensional compressible \(N - S\) equation and a fully buoyant corrected Renormalization-group (RNG) \(k - \varepsilon\) turbulence model to build a fire smoke spread model. Additionally, a cellular automaton model is employed to construct a simulation model for the evacuation of personnel in interval tunnels. We used the models to investigate the influence of longitudinal ventilation speed on smoke spread and the evacuation behavior characteristics of personnel under a moving subway train with fire. Results show that smoke spreads downstream of the fire source, and the temperature of smoke in tunnels decreases as longitudinal ventilation speed increases. A prediction model between longitudinal ventilation and the peak value of smoke temperatures in tunnels was modified based on Li's prediction model. Meanwhile, the total evacuation time decreases as the longitudinal ventilation speed increases. A theoretical prediction model between the peak value of smoke temperatures and total evacuation time is developed. The model parameters are determined using a nonlinear fitting method. The influence of longitudinal ventilation on the average flow rate and arrival time at the exit upstream of the fire source is less. However, it has a significant effect downstream of the fire source. As the longitudinal ventilation speed increases, the average flow rate at the exit downstream of the fire source increases, leading to a decrease in total evacuation time. A notable consideration is that the elderly or minors are significantly affected by smoke in the late stages of evacuation process, leading to an increase in total evacuation time.
