Impact of turbulence on the product, modelling and simulation

先前的研究表明，在所有大型纳米颗粒生产反应器中发生的湍流对产品性能有巨大的影响：例如，聚集体的尺寸增长可以通过快速冷却来限制，而快速冷却只能通过快速湍流混合冷惰性气体来实现。然而，同时，如果这些产品经历一个很大程度上随机的温度和浓度历史，则产品性能可以变得更加均匀。不幸的是，湍流对颗粒影响的预测和模拟并非微不足道，因为湍流混合会导致颗粒浓度场在非常小的尺度上的不均匀性，即使有巨大的计算能力也无法充分解决。因此，人们使用时间平均模拟（RANS），或者在最近的工作中使用空间滤波模拟（LES），它们允许可接受的计算时间。不幸的是，平均或过滤意味着不可挽回的信息损失，从而导致错误的预测，除非以足够的精度估计丢失的信息（使用适当的模型）。因此，在“Forschergruppe”的第三阶段，将进行实验，一方面研究湍流对产品性能的影响，另一方面以非常高的分辨率确定最小的空间尺度（在小空间域中）。本项目将使用最先进的LES方法和模型重新计算已完成的实验。这样，实验就可以被解释，发生的现象也可以被分析。与此同时，相关的、可用的和开发的湍流模型将被测试、改进，如果可能的话，在基于新的、详细的测量实验的模拟中得到验证。这些经过验证的模拟结果将显示湍流对产品特性（如粒度分布或团聚程度）的影响，并进一步解释它们。

Previous work has shown that turbulence, which occurs in all large nano-particle production reactors, has a massive influence on product properties: for example, the size growth of aggregates can be limited by rapid cooling, which can only be achieved by rapid and thus turbulent mixing of cold, inert gas. At the same time, however, product properties can be made more uniform if these products undergo a - largely - random temperature and concentration history. Unfortunately, the prediction and simulation of the turbulence effects on the particles is not trivial, since turbulent mixing causes inhomogeneity of the particle concentration fields on very small scales, which cannot be sufficiently resolved even with massive computing power. One therefore uses time-averaged simulations (RANS) or, in recent work, spatially filtered simulations (LES), which allow acceptable computational time. Unfortunately, averaging or filtering implies an irretrievable loss of information that leads to erroneous predictions unless the lost information is estimated with sufficient accuracy (using appropriate models). Therefore, in the third phase of the „Forschergruppe", experiments will be performed to investigate the effect of turbulence on product properties on the one hand, and to determine the smallest spatial scales (in a small spatial domain) with very high resolution on the other hand. The present project will recalculate the performed experiments using state-of-the-art LES methods and models. In this way, the experiments can be interpreted and phenomena that occur can be analyzed. At the same time, the relevant, available and developed turbulence models will be tested, improved, and, if possible, validated in the simulation based on the new, detailed measured experiments. The results of these validated simulations will then show the influence of turbulence on product properties such as particle size distribution or degree of agglomeration, and moreover, explain them.