Application of a Stochastic Reactor Model Approach for Prediction of Gas Turbine Engine Emissions

西门子加拿大有限公司（SCL）是一家生产用于发电的工业航空衍生燃气轮机的制造商。鉴于SCL不断努力开发符合最严格排放标准的燃烧系统，因此需要不断改进创新的低排放概念设计，并迅速推向市场。在设计过程的早期阶段，使用简化模型来了解任何改变对排放水平的全球影响。与复杂的模型或实验相比，这种对简化建模的依赖降低了成本；然而，在这些简化模型中存在很大程度的不确定性。了解不确定性将有助于更好地决定在开发过程中进一步追求什么设计，避免因超过预期的开发时间而增加成本。简化模型中最大的不确定性来源之一是无法以足够的保真度捕获湍流-化学相互作用（TCI），以确保可靠的分析。开发一种能够捕获TCI的计算效率高的燃气轮机简化建模技术，将降低开发新发动机设计所需的成本和时间，提高SCL的竞争力。目前，这种简化的燃气轮机建模技术还没有达到与CFD等高端商用仿真工具相同的成熟度。拟议的研究将应用现有的随机反应堆模型（SRM）软件来模拟燃气轮机排放，以解决SCL排放建模的挑战。该项目的重点是“概念验证”，证明SRM方法不仅对燃气轮机排放建模可行，而且还提供了一种强大且成本效益高的方法。SRM方法的独特之处在于它可以捕获TCI，同时具有较低的计算成本。主要里程碑和可交付成果是：(1)将SRM方法应用于燃气轮机的代码修改，(2)确定燃气轮机建模的初始模型参数，(3)使用SRM方法对燃气轮机排放进行初步概念模拟验证，(4)使用实验数据评估SRM方法，以及(5)关于概念验证研究结果的报告。

Siemens Canada Limited (SCL) is a manufacturer of industrial aero-derivative gas turbines for power generation. Given ever increasing effort by SCL to develop combustion that meet the most stringent emissions standards, there is a need to continually improve how new innovative low emission concepts can be design and brought to market rapidly. At the early stages of the design process, simplified models are utilized to understand global implications on emission levels of any alterations. This reliance on simplified modeling reduces the cost versus complex models or experimentation; however, there are large degrees of uncertainty in these simplified models. Understanding the uncertainty would allow for better decisions as to what designs are pursued further in the development process, avoiding costs increased from longer than expected development time. One of the largest sources of uncertainties in the simplified models is the inability to capture turbulence-chemistry interaction (TCI) with adequate fidelity to ensure reliable analysis. Developing of a computationally-efficient simplified modeling technique for gas turbines that can capture TCI will reduce the cost of and time required for developing new engine designs, enhancing the competitiveness of SCL. Currently, such simplified modeling techniques for gas turbines have not achieve the same level of maturity as the high end commercially available simulation tools like CFD. The proposed research will apply an existing Stochastic Reactor Model (SRM) software to modeling gas turbine emissions to address SCL emissions modeling challenges. The project is focused on a "proof of concept", demonstrating that a SRM approach is not only feasible for modeling gas turbine emissions, but also provide a robust and cost effective approach. The SRM approach is unique in that it can capture TCI while having low computational costs.The major milestones and deliverables are: (1) code modifications to apply the SRM approach to gas turbines, (2) determination of initial model parameters for gas turbine modeling, (3) initial proof of concept simulation of gas turbine emissions using the SRM approach, (4) assessment of the SRM approach using experimental data, and (5) a report on the findings of the proof of concept study.