Efficient In-Situ and Reduced Chemistry Algorithms for Chemical Process Flow Simulation

该项目的主要目标是开发和实施高效的原位和简化的化学算法，用于详细的化学过程流模拟。目前化工过程工业流动模拟的趋势是应用计算流体动力学（CFD）来更完整地描述化学反应器内的流场和化学物质分布。对于化学反应流动，处理由复杂动力学方案（即大于10到20种）产生的化学源项通常是流动模拟中计算最密集的部分。因此，通过利用复杂动力学方案的两个共同特征，即化学时间尺度上的大分离（相对于流动时间尺度）和“访问区域”的紧密性（即，在特定模拟中实际发生的成分），计算效率的显著提高是可能的。我们将研究利用这些特性的两种方法。第一个是原位自适应制表（ISAT），适用于中等规模的动力学方案（大约20种化学物质），由于其实现的灵活性，应该是强大的化学过程流模拟。第二种，内在低维流形（ILDM）方法，适用于更大的动力学方案（前提是大多数化学时间尺度小于流动时间尺度），但计算上更难以实现，因为它需要在每次改变动力学方案或流动参数后重新计算。为了克服每个单独算法的缺点，将探索ILDM-ISAT组合算法的实现。这些算法将在现有的全概率密度函数（PDF）代码中实现，该代码是为紊流反应流模拟而设计的。将进行两种工业化学反应方案（甲烷热氯化和乙烯自由基聚合）的完整pdf模拟，以便在广泛的实际工厂反应器条件下验证和记录算法的效率。

The principal objectives of this project are the development and implementation of efficient in-situ and reduced chemistry algorithms for detailed chemical process flow simulation. The current trend in flow simulation in the chemical process industries is towards the application of computational fluid dynamics (CFD) to obtain a more complete descriptions of the flow field and chemical species distribution inside chemical reactors. For chemically reacting flows, treatment of the chemical source terms resulting from complex kinetic schemes (i.e. greater than 10 to 20 species) is often the most computationally intensive part of a flow simulation. Considerable gains in computational efficiency are thus possible by exploiting two common features of complex kinetic schemes, namely, the large separation in chemical time scales (relative to flow time scales), and the compactness of the "accessed region" (i.e., the compositions that actually occur in a particular simulation). Two methodologies that make use of these features will be investigated. The first, in-situ adaptive tabulation (ISAT), is applicable to moderately large kinetic schemes (around 20 chemical species) and, due to its flexibility of implementation, should be powerful for chemical process flow simulations. The second, intrinsic low-dimensional manifold (ILDM) method, is applicable to larger kinetic schemes (provided that the majority of the chemical time scales are smaller than the flow time scales), but computationally more difficult to implement since it requires retabulation after each change in the kinetic scheme or flow parameters. In order to overcome the shortcomings of each individual algorithm, implementations of a combined ILDM-ISAT algorithm will be explored. The algorithms will be implemented in an existing full probability density function (PDF) code designed for turbulent reacting flow simulation. Full pdf simulations of two industrial chemical reaction schemes (thermochlorination of methane and free-radical poly merization of ethylene) will be carried out in order to validate and document the efficiency of the algorithms under a wide range of actual plant reactor conditions.