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

这个项目是为复杂和湍流的多尺度流动问题开发新一代预测方法。pi正在寻找将整个流域自然分离成多个区域，每个区域都有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实验和计算技术。我们的目标是开发通用的面向数据技术，允许跨不同区域的不同技术之间的无缝接口。数据处理涉及到不同区域之间的流动物理匹配，以及数值数据的一致传输。所采用的方法包括两种技术：湍流大涡模拟（LES）和多维测速实验，特别是粒子图像测速（PIV）。并行PIV实验和LES计算之间的动态数据处理方法是使用条件统计，近似于随机估计过程。上述实验与计算相结合的动态数据面向概念将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，是广泛工程应用的基础。在许多化学工业处理装置中，实现混合的速度决定了工厂的总生产量和盈利能力。在其他设备中，例如依靠流体混合物组合的涡轮机，热效率以一种敏感的方式取决于混合的质量和速度。放大对好的工厂设计至关重要。通过综合PIV和PLIF实验和大涡模拟，将使用一系列实验室到中试工厂规模的搅拌槽几何形状来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.