Fluidized particulate systems can be well described by coupling the discrete element method (DEM) with computational fluid dynamics (CFD). However, the simulations are computationally very demanding. The computational demand is drastically reduced by applying the coarse grain (CG) approach, where several particles are summarized into larger grains. Scaling rules are applied to the dominant forces to obtain precise solutions. However, with growing grain size, an adequate representation of the interaction forces and, thus, representation of sub-grid effects such as bubble and cluster formation in the fluidized particulate system becomes challenging. As a result, particle drag can be overestimated, leading to an increase in average particle height. In this work, limitations of the system-to-grain ratio are identified but also a dependency on system width. To address this issue, sub-grid drag models are often applied to increase the accuracy of simulations. Nonetheless, the sub-grid models tend to have an ad hoc fitting, and thorough testing of the system configurations is often missing. Here, five different sub-grid drag models are compared and tested on fluidized bed systems with different Geldart group particles, fluidization velocity, and system-to-grain diameter ratios.

通过将离散元法 (DEM) 与计算流体动力学 (CFD) 相结合，可以很好地描述流化颗粒系统。然而，模拟对计算的要求非常高。通过应用粗粒 (CG) 方法，将多个粒子汇总为更大的粒子，计算需求大大减少。将缩放规则应用于主导力以获得精确的解。然而，随着颗粒尺寸的增大，相互作用力的充分表示以及流化颗粒系统中气泡和团簇形成等亚网格效应的表示变得具有挑战性。因此，颗粒阻力可能被高估，导致平均颗粒高度增加。在这项工作中，确定了系统与颗粒比率的限制，但也确定了对系统宽度的依赖性。为了解决这个问题，通常应用子网格阻力模型来提高模拟的准确性。尽管如此，子网格模型往往具有临时拟合，并且经常缺少对系统配置的彻底测试。在这里，在具有不同 Geldart 组颗粒、流化速度和系统与颗粒直径比的流化床系统上对五种不同的子网格阻力模型进行了比较和测试。