Ryder Noah L, Sutula Jason A, Schemel Christopher F, Hamer Andrew J, Van Brunt Vincent
Packer Engineering Inc., 6700 Alexander Bell Drive, Suite 100, Columbia, MD 21046, USA.
J Hazard Mater. 2004 Nov 11;115(1-3):149-54. doi: 10.1016/j.jhazmat.2004.06.018.
The use of Computational Fluid Dynamics (CFD) and in particular Large Eddy Simulation (LES) codes to model fires provides an efficient tool for the prediction of large-scale effects that include plume characteristics, combustion product dispersion, and heat effects to adjacent objects. This paper illustrates the strengths of the Fire Dynamics Simulator (FDS), an LES code developed by the National Institute of Standards and Technology (NIST), through several small and large-scale validation runs and process safety applications. The paper presents two fire experiments--a small room fire and a large (15 m diameter) pool fire. The model results are compared to experimental data and demonstrate good agreement between the models and data. The validation work is then extended to demonstrate applicability to process safety concerns by detailing a model of a tank farm fire and a model of the ignition of a gaseous fuel in a confined space. In this simulation, a room was filled with propane, given time to disperse, and was then ignited. The model yields accurate results of the dispersion of the gas throughout the space. This information can be used to determine flammability and explosive limits in a space and can be used in subsequent models to determine the pressure and temperature waves that would result from an explosion. The model dispersion results were compared to an experiment performed by Factory Mutual. Using the above examples, this paper will demonstrate that FDS is ideally suited to build realistic models of process geometries in which large scale explosion and fire failure risks can be evaluated with several distinct advantages over more traditional CFD codes. Namely transient solutions to fire and explosion growth can be produced with less sophisticated hardware (lower cost) than needed for traditional CFD codes (PC type computer verses UNIX workstation) and can be solved for longer time histories (on the order of hundreds of seconds of computed time) with minimal computer resources and length of model run. Additionally results that are produced can be analyzed, viewed, and tabulated during and following a model run within a PC environment. There are some tradeoffs, however, as rapid computations in PC's may require a sacrifice in the grid resolution or in the sub-grid modeling, depending on the size of the geometry modeled.
使用计算流体动力学（CFD），特别是大涡模拟（LES）代码来模拟火灾，为预测大规模效应提供了一种有效工具，这些效应包括羽流特性、燃烧产物扩散以及对相邻物体的热效应。本文通过几个小规模和大规模的验证运行以及过程安全应用，阐述了由美国国家标准与技术研究院（NIST）开发的大涡模拟代码——火灾动力学模拟器（FDS）的优势。本文介绍了两个火灾实验——一个小房间火灾和一个大型（直径15米）池火。将模型结果与实验数据进行了比较，结果表明模型与数据之间具有良好的一致性。然后，通过详细介绍一个罐区火灾模型和一个受限空间内气体燃料点火模型，将验证工作扩展到证明其对过程安全问题的适用性。在该模拟中，一个房间充满丙烷，经过一段时间扩散后被点燃。该模型得出了气体在整个空间内扩散的准确结果。这些信息可用于确定空间中的可燃极限和爆炸极限，并可用于后续模型中，以确定爆炸产生的压力波和温度波。将模型扩散结果与工厂互助保险公司进行的一项实验进行了比较。通过上述示例，本文将证明FDS非常适合构建实际的过程几何模型，在该模型中，可以评估大规模爆炸和火灾失效风险，与更传统的CFD代码相比具有几个明显优势。具体而言，与传统CFD代码（PC型计算机与UNIX工作站）相比，使用不太复杂的硬件（成本更低）就能生成火灾和爆炸发展的瞬态解，并且可以用最少的计算机资源和模型运行时长求解更长的时间历程（计算时间约为数百秒）。此外，在PC环境下，模型运行期间及之后，可以对生成的结果进行分析、查看和制表。然而，也存在一些权衡，因为在PC上进行快速计算可能需要在网格分辨率或亚网格建模方面做出牺牲，这取决于所建模几何形状的大小。
