Probabilistic Model of Smoke Filling in Large Spaces

There are many uncertain factors to affect the process of smoke filling in large spaces. In this paper, experimental studies on natural smoke filling induced by an electric car fire were carried out to compare with the model of smoke filling in large spaces. The experimental results are in good agreement with the theoretical values. Afterwards, Latin Hypercube Sampling method was used to mathematically describe the probabilistic model of smoke filling in large spaces. Latin Hypercube Sampling method was used to carry out a computer simulation sampling of the uncertain factors in the process of smoke filling in large space. The probability density function was used to describe the randomness of the uncertain factors in the smoke filling process. Then, the Pathfinder simulation and data review, the time needed for evacuation was obtained. By comparing the difference between the time of smoke filling and evacuation, the failure probability of smoke control under the effect of uncertain factors was obtained. The probability model developed in this paper to determine the failure progress is more reasonable than the traditional method. The model is more suitable to assess the risk of large spaces building fire and can also combine with the risk tolerance to develop safety assessment.