细粉下料过程的气固流体动力学作用分析

细颗粒粉体下料时受气固流体力学作用在料仓出口附近形成逆压力梯度，使得粉体下料流率实验值远低于理论预测值。而且该压力梯度力直接测量较困难，对模型修正和发展提出了挑战。以玻璃微珠、流化催化裂化（FCC）催化剂颗粒、褐煤和聚氯乙烯（PVC）颗粒为实验材料，首先开展粉体静力学与动力学测试，借助休止角（AOR）、豪斯纳比（HR）和卡尔流动指数（CFI）多个粉体流动性判据综合分析不同粉体的流动特性；在分析粉体料仓出口附近气固流动特征的基础上，结合Jenike流动与不流动判据，将作用在细颗粒粉体上的逆压力梯度力引入到拱应力平衡方程；进一步，提出了利用迭代算法获得逆压力梯度力的方法，实现了对逆压力梯度力与粉体料仓下料流率的预测。建立的粉体下料流率模型考虑了气固流体动力学作用对粉体下料流动的影响，有效改善了传统模型对细粉体流率预测偏高的问题，模型预测偏差从60%以上降低至±20%。

Analysis of the gas-solid fluid dynamic interaction on fine powder discharge

When the fine-particle powder is blanked, it is subjected to gas-solid hydrodynamics to form an inverse pressure gradient near the outlet of the silo, making the experimental value of the powder blanking flow rate far lower than the theoretical prediction value. Moreover, this inverse pressure gradient force is difficult to measure directly, posing a challenge to model revision and development. In this paper, glass beads, fluidized catalytic cracking (FCC) catalyst particles, lignite and PVC particles were used as experimental materials. Firstly, the static and kinetic tests of powders were carried out, and the flow characteristics of different powders were analyzed by angle of repose (AOR), Hausner ratio (HR) and Carr flowability index (CFI). Immediately afterwards, based on the analysis of the gas-solid flow characteristics near the outlet of the powder hopper, and combined with the Jenike flow and non-flow criterion, the reverse pressure gradient force acting on the fine powder was introduced into the arch stress equilibrium equation. Furthermore, a logical block diagram of the algorithm for obtaining the inverse pressure gradient force using an iterative algorithm is proposed to achieve the prediction of the inverse pressure gradient and the modification of the silo discharging model. The established powder discharging flow rate model takes into account the influence of gas-solid hydrodynamics on the powder discharging flow, which effectively improves the disadvantage of the traditional model in predicting the fine powder flow rate, and the deviation of the model prediction is reduced from over 60% to ±20%.