Contributions to Building Smoke Management

加拿大每年平均发生火灾约6万起，造成3,300多人伤亡，损失12亿美元。吸入烟雾或有毒气体，而不是火烧伤，被认为是建筑物火灾中死亡的主要原因。确定火灾时烟气的运动规律，设计可靠的烟气管理系统，对于防止火灾时烟气堵塞疏散通道具有重要意义，尤其是对于高层建筑和大空间建筑，人员暴露在烟气中的风险更高。计算机模型是高层建筑烟雾管理不可或缺的工具，有三种类型的计算机模型可用于不同的精度和计算成本：网络，区域和计算流体动力学（CFD）模型。目前，没有一个单独使用的这些模型可以令人满意地执行快速的整体建筑分析，并同时提供所需的精度和详细的空间信息的大型或近火空间。因此，本研究的目的是发展一个公共领域的综合计算机模型，它集成了网络和CFD火灾模型，大大提高整个建筑物的烟雾运动分析。集成模型将在需要详细设计信息的空间使用CFD，并在建筑物的其余部分使用网络模型进行快速计算。本研究将增加一个能量方程解算器，以使网路模型适合于火灾烟雾传输的模拟。一个最好的集成策略将被确定，以解决两个模型的时间不一致性。综合模型将验证对现有的全尺寸的数据在文献中，并在一个小规模的测试设备的实验室实验，并进一步应用到实际的高层和摩天大楼建筑。该研究还将导致更好地了解建筑烟雾传播的物理学和对烟雾控制系统性能的见解。该研究将有助于设计更安全的建筑物，减少生命和财产损失，并培训HQP为他们未来在加拿大建筑烟雾管理和消防中的作用。

The average annual fire incidents in Canada were around 60,000 with over 3,300 injuries and deaths, and a 1.2 billion dollar loss. Inhalation of smoke or toxic gases, rather than fire burns, is considered the major cause of fatality in building fires. It is of great importance to determine smoke movement and design a reliable smoke management system to prevent smoke from blocking evacuation route during a fire, especially for high-rise buildings and buildings with large spaces, where there is a higher risk that occupants will be exposed to smoke. Computer models are indispensable tools for high-rise building smoke management and three types of computer models are available at varying accuracy and computational cost: network, zone, and computational fluid dynamics (CFD) models. Presently, none of these models used individually can satisfactorily perform a fast whole-building analysis and simultaneously provide required accuracy and detailed spatial information for large or near-fire spaces. Therefore, this research aims to develop a public-domain integrated computer model, which integrates a network and a CFD fire model, to greatly improve whole-building analyses of smoke movement. The integrated model will use the CFD for spaces where detailed design information is required, and the network model for the rest of a building for fast computation. This study will add an energy equation solver to suit the network model to the simulations of fire smoke transport. A best integration strategy will be identified to resolve the temporal inconsistency of the two models. The integrated model will be validated against the existent full-scale data in the literature and by the laboratory experiments at a sub-scale test facility, and further applied to actual high-rise and skyscraper buildings. The research will also result in a better understanding of the physics of building smoke propagation and insights into the performance of smoke control systems. The research will contribute to designing safer buildings, reducing life and property losses, and training of HQP for their future roles in building smoke management and fire protection in Canada.