GOALI: Integrated Experimental and Computational Multi-Zonal Approach to Multiple-Scale Problems: Flow in a Stirred Tank Reactor

该项目旨在开发新一代预测方法，用于复杂和湍流的多尺度流动问题。PI正在寻找将整个流域自然分离成多个区域，每个区域的特征是其自己的长度和时间尺度范围。将酌情在不同的领域采用不同的技术，包括实验和计算技术。其目标是开发通用的面向数据的技术，允许跨不同区域的不同技术之间的无缝接口。面向数据涉及跨不同区域之间的界面的流动物理匹配，以及数值数据的一致传输。所采用的方法包括两种技术：湍流的大涡模拟(LES)和多维测速实验，特别是粒子图像测速(PIV)。在并行PIV实验和大涡模拟计算之间的动态数据面向的方法是使用条件统计量，通过随机估计过程进行近似，将上述集成实验和计算之间的动态数据面向的概念应用于搅拌槽内的流动问题。两种或两种以上流体的混合是一个复杂的物理过程，是广泛的工程应用的基础。在许多化学工业处理装置中，混合的速度决定了装置的总体生产能力和盈利能力。在其他设备中，例如依靠混合流体的涡轮机，热效率以一种敏感的方式取决于混合的质量和速度。扩大规模是好的工厂设计的关键。通过集成的PIV和PLIF实验和大涡模拟，将使用一系列实验室来试验工厂规模的搅拌槽几何形状，以解决宏观尺度混合的准确预测和可靠的放大问题。

Abstract - Balachandar - 9910543This project is to develop a new generation of predictive methodology for complex and turbulent multiple-scale flow problems. The PIs are looking for natural separation of the entire flow domain into multiple zones, each characterized by their own range of length and time scales. Different techniques, both experimental and computational, will be employed over the different domains, as appropriate. The goal is the development of generalized data-facing that will permit seamless interfaces between the different techniques across different zones. Data-facing involves matching of flow physics across interfaces between different zones, as well as consistent transfer of numerical data. The methodology to be used involves two techniques: large eddy simulation (LES) of turbulent flow and multi-dimensional velocimetry experiments, in particular, particle image velocimetry (PIV). The approach to dynamic data-facing between parallel PIV experiments and LES computations is the use of conditional statistics, approximated by the stochastic estimation procedure.The above concept of dynamic data-facing between integrated experiments and computations will be applied to the problem of flow in a stirred tank. The mixing of two or more fluids is a complex physical process that is fundamental to a wide range of engineering applications. In many chemical industrial processing units the rate at which mixing can be achieved determines the overall throughput and profitability of the plant. In other devices, such as turbines that rely on the combination of fluid mixtures, the thermal efficiency depends on the quality and speed of mixing in a sensitive way. Scale-up is critical to good plant design. Through integrated PIV and PLIF experiments and large eddy simulations, accurate prediction of macro-scale mixing and the problem of reliable scale-up will be addressed using a range of laboratory to pilot plant scale stirred tank geometries.