The effect of particle clusters on the chemical performance of a downer reactor during fluid catalytic cracking of gasoil is investigated by performing several 4-way coupled reactive 3D Discrete Element Method - Large Eddy Simulations (DEM-LES) in OpenFOAM. A one-meter-long cylindrical downer was used as a canonical case to analyze the dynamics of particles via 3D DEM-LES. A fluid mesh of 7 × 10⁶ hexahedral cells sufficed to meet the $y^{+}<1$ criterion for wall-resolved simulations. A four-lump kinetic model was selected to describe the chemistry of the process in which gasoil, gasoline, light gases, and coke are considered as the lumps. The difference on the validation of the reactive 3D DEM-LES with experimental data found in literature was 10 rel.% at a catalyst-to-oil ratio of 8.1 and less than 5 rel.% at a catalyst-to-oil ratio of 15 and higher. Two effectiveness factors that take into account the frequency, mean lifetime, size and shape of clusters were proposed to capture the effects of clusters on the prevailing chemistry. In this way, it is possible to translate the information obtained with the intensive 3D DEM-LES to a regular 1D model. The first effectiveness factor was based on two empirical correlations that relate the particle Reynolds number and the Stokes number with the frequency and mean lifetime of clusters, determined from 3D DEM-LES. A second effectiveness factor relates to the shape and size of clusters to compute the efficiency of the catalytic reactions. A modified 1D model that takes into account these two effectiveness factors predicted values of conversion within 3 rel.% of deviation from experimental data, similar to those of the 3D DEM-LES with a computational time 4 × 10⁸ times lower.

通过在OpenFOAM中执行几种4向耦合反应3D离散元方法-大涡模拟（DEM-LES），研究了瓦斯油流化催化裂化过程中颗粒簇对唐纳反应器化学性能的影响。使用一米长的圆柱状下降器作为典型案例，通过3D DEM-LES分析粒子的动力学。一个7  ×  10 ^6个六面体单元的流体网格足以满足${ÿ}^{+}<\mathrm{1个}$墙解析模拟的标准。选择四集总动力学模型来描述过程化学，其中将瓦斯油，汽油，轻质气体和焦炭视为集总。在文献中发现的实验数据对反应性3D DEM-LES的验证差异在催化剂与油的比为8.1时为10 rel。％，在催化剂与油的比为15时小于5 rel。％和更高。提出了两个有效因素，考虑了频率，平均寿命，团簇的大小和形状，以捕获团簇对主要化学物质的影响。这样，可以将通过密集3D DEM-LES获得的信息转换为常规1D模型。第一个有效性因素基于两个经验相关性，这两个相关性将粒子雷诺数和斯托克斯数与从3D DEM-LES确定的团簇的频率和平均寿命相关联。第二个有效性因素与簇的形状和大小有关，以计算催化反应的效率。修改后的一维模型考虑了这两个有效性因素，预测的转换值与实验数据之间的偏差在3 rel。％之内，类似于3D DEM-LES的计算时间为4 ×低 10 ⁸倍。