Systematic Evaluation and Development of Discrete Element Coarse- Graining Approaches for the Modeling of Fluidized Particle Systems

计算流体力学(CFD)与粗粒度离散元方法(CFD-CGDEM)的耦合是一种比较新的、高效的颗粒系统数值模拟方法。基于离散单元法(DEM)，将选定数量的粒子组合成一个具有代表性的地块。这大大减少了需要跟踪的粒子数量，并导致了极大的加速。因此，它是一种在工业规模上模拟颗粒系统的有前途的方法。与经典的DEM一样，通过求解牛顿运动定律来计算每个地块的轨迹，并使用适当的模型来描述接触和流体-颗粒相互作用。由于能量守恒或基于相似理论，颗粒-颗粒、颗粒-壁面和颗粒-流体相互作用力需要进行定标。在过去的几年里，许多作者提出了几种扩展方法，但尚未对这些方法进行系统的比较。所要求的项目将首先对不同的扩展方法进行基本和系统的比较。目的是找到一种在各种比例因子上都能给出可靠结果的比例方法。初步研究表明，特征反应器长度与颗粒直径之比不应超过一定限度。将开发一种自适应粗粒化方法，以允许反应堆内窄区域和宽区域的不同比例因子。将研究这种方法是否允许CFD-CGDEM有效地用于内部部件紧密堆积的反应堆，这些内部部件在可能的技术应用中特别相关。最初的CFD-DEM要求是细胞比颗粒大。因此，粗粒化会导致网格固有的粗化。这就是为什么无法解决重要的中尺度效应(例如，集群形成)的原因。将调查覆盖此单元大小要求的方法是否会导致CFD-CGDEM更高的精度。此外，还将研究使用过滤阻力模式模拟中尺度效应是否对结果质量有积极影响。到目前为止，关于CFD-CGDEM在多分散颗粒系统中的适用性的研究还很少，而且非常基础。正确预测颗粒偏析的重要方面尚未得到研究。该项目的目的是缩小这一差距，并评估CFD-CGDEM在原则上是否可以用于预测双分散颗粒系统的分离行为。基于所开发的方法，将使用CFD-CGDEM对包括内件在内的工业相关案例进行模拟。将对单分散和双分散系统进行研究。目的是证明CFD-CGDEM是一种有效和可靠的方法，用于模拟“真实世界”应用中的颗粒系统，并确定其应用范围。

The coupling of Computational Fluid Dynamics (CFD) and a Coarse- Grained Discrete Element Method (CFD-CGDEM) is a relatively new and efficient numerical method for the simulation of particulate systems. Based on the Discrete Element Method (DEM) a chosen number of particles is combined into one representative parcel. This reduces the number of particles that need to be tracked significantly and leads to an enormous speed-up. Therefore, it is a promising method for the simulation of particulate systems on an industrial scale. Like in classical DEM the trajectories of each parcel are calculated by solving Newton’s law of motion and appropriate models are used to describe contact and fluid-particle interactions. Due to the conservation of energy or based on similarity theory the particle- particle, particle-wall, and particle-fluid interaction forces need to be scaled. In the last years, several scaling approaches by numerous authors were proposed, but a systematic comparison of those approaches has yet not been done. The requested project will start with a fundamental and systematic comparison of different scaling approaches. Aim is to find the scaling approach that gives reliable results over a wide range of scaling factors. Preliminary studies have shown that the ratio of characteristic reactor length to parcel diameter should not exceed a certain limit. An adaptive Coarse-Graining- Method will be developed to allow different scaling factors for narrow and wide regions within the reactor. It will be investigated if this approach allows CFD-CGDEM to be efficiently used for reactors with closely packed internals which are especially relevant in possible technical applications. An original CFD-DEM requirement is that the cells are bigger than the particles. Therefore, Coarse-Graining leads to an inherent coarsening of the mesh. That is why important meso-scale effects (e.g., cluster formation) cannot be resolved. It will be investigated if methods that override this cell size requirement lead to a higher accuracy of CFD-CGDEM. Furthermore, it will be examined if a modelling of meso-scale effects using filtered drag models has a positive effect on the quality of the results. By now only few and very basic studies have been done on the usability of CFD-CGDEM for poly-disperse particulate systems. The important aspect of a correct prediction of particle segregation has not yet been investigated. Aim of the project is to close this gap and to evaluate if CFD-CGDEM can in principle be used to predict the segregation behavior of bi- dispersed particulate systems. Based on the developed methods CFD-CGDEM will be used for the simulation of an industrially relevant case including internals. Mono- and bi-dispersed systems will be investi-gated. Aim is to prove that CFD-CGDEM is an efficient and reliable method for the simulation of particulate systems for “real world”-applications as well as to identify application limits.