Research on fire scenario analysis and emergency response strategies for L-shaped buildings using FDS

With the rapid development of information and communication technologies, modern fire rescue models are evolving towards informatization, proactivity, and spatialization. Existing fire simulation models primarily focus on the impact of individual building parameters, lacking a systematic analysis of the multi-factor coupling effects on fire spread. As a result, it is difficult to accurately predict the trend of fire spread and meet the demands of dynamic escape routes for three-dimensional rescue systems. In this study, a typical L-shaped building was used as the research object, and a parameterized fire scenario database was constructed using the FDS (Fire Dynamics Simulator) fire simulation software, combined with experimental data for multi-factor coupling analysis. By simulating various factors such as initial fire locations (stairwell or corridor corner), floor height of ignition (low/mid/high area), and wind direction, this study systematically reveals the dynamic evolution process of smoke diffusion and the spatial distribution characteristics and interaction mechanisms of fire smoke in indoor and outdoor three-dimensional spaces. It also focuses on the time-dependent variations of smoke propagation in indoor escape routes, staircases, and external walls. The results indicate that the L-shaped building causes asymmetric smoke diffusion. The fire spread mode is a result of multi-factor coupling effects at the fire scene. Safe evacuation decisions are directly related to fire spread characteristics, smoke diffusion paths, the location of trapped individuals, fire alarm occurrence time, and rescue methods. Minor changes in each parameter can lead to significant differences in evacuation outcomes. Based on the situation at the site, generating real-time, dynamic escape paths is a crucial guarantee for improving rescue efficiency and success rates. The findings provide scientific support for the design and optimization of human-machine collaborative rescue systems and offer a data foundation for fire risk assessment and emergency response planning.