Development of tools for the numerical simulation of turbulent combustion

碳氢化合物的燃烧是许多工程应用中的重要现象。完整描述这些体系所需的化学动力学机制通常非常大，通常包括涉及数百个化学物种的数千个反应。在几乎所有燃烧过程的实际应用中，发生燃烧的流动是湍流的，化学过程和流动过程之间存在复杂的、非线性的相互作用。 在使用随机过程模拟燃烧过程方面的新进展已经开始为了解化学反应等过程的离散性质提供了新的见解。最近的一个随机粒子模型(SPM)被用来研究离散反应对点火延迟时间的逐个变化的影响。了解这些变化对于设计和运行先进的新发动机概念(如HCCI发动机)至关重要。我们将使用SPM在两种情况下研究点火： 1-点火同质性，通常假设整个均质混合物同时点火，但有证据表明，可能存在一个在导致火焰传播的点上点火的区域；以及 2-使用随机粒子模型的大涡模拟湍流射流的点火将被用来产生独立的简化化学流形，这将使我们能够研究湍流和化学反应变化的独立影响。 我们还将使用线性涡流模型，这是一种在反应流动的一维模拟中近似湍流混合影响的模型，以改进对湍流预混燃烧中反应进程变量的概率密度函数的预测，以期改进我们对湍流预混火焰的预测。

Combustion of hydrocarbons is an important phenomenon in many engineering applications. The chemical kinetic mechanisms required to completely describe these systems are usually extremely large, often comprising thousands of reactions involving hundreds of chemical species. In virtually all practical applications of combustion processes, the flow in which the combustion takes place is turbulent and there is a complex, non-linear interaction between the chemical and flow processes. New developments in the use of stochastic processes to simulate combustion processes have begun to offer new insight into the discrete nature of processes like chemical reactions. A recent stochastic particle model (SPM) has been used to study the effects of discrete reactions on the shot-to-shot variations in ignition delay times. Understanding these variations is critical to the design and operation of advanced new engine concepts such as HCCI engines. We will use the SPM to study ignition in two contexts: 1 - ignition homogeneity, where it is often assumed that the entire homogeneous mixture ignites simultaneously, but evidence suggests that there may be a regime in which ignition occurs in spots that lead to propagating flames; and 2 - ignition of turbulent jets using Large Eddy Simulation with the stochastic particle model will be used to generate independent reduced chemistry manifolds that will allow us to study the independent effects of turbulence and chemical reaction variations. We will also use the Linear Eddy Model, a model that approximates the impact of turbulent mixing in one-dimensional simulations of reacting flows, to generate improved predictions of the probability density function of reaction progress variable in turbulent premixed combustion, with a view to improving our predictions of turbulent premixed flames.