基于Voronoi-Random算法微观结构褶型过滤介质过滤性能

Filtration performance of microstructure pleated filter media based on Voronoi-Random algorithm

在利用数值模拟研究褶型纤维过滤介质过滤性能时,为了克服层状结构纤维层与层之间不连贯及孔状结构纤维规整分布的不足,本文基于Voronoi-Random算法建立了褶型纤维过滤介质模型,并通过加密或减少纤维数量改变其固体体积分数(SVF).采用“Caught on first touch”和“Hamker”两种碰撞模型对褶型纤维过滤介质的气-固两相流动进行数值模拟.结果表明:压力损失和过滤效率数值计算值与经验关联式计算值吻合较好,误差在15%之内;将采用两种碰撞模型得到的过滤效率与经验关联式计算值进行比较,得出采用“Hamker”碰撞模型得到的过滤效率符合实际;在过滤介质上的沉积颗粒并不完全是纤维捕集,大多数颗粒是由形成的树枝状结构捕集;压降和单位面积沉积量随过滤时间的增长呈现指数增长趋势.

When using numerical simulation to study the filtration performance of pleated fiber filter media,in order to overcome the shortcomings of discontinuous layers and regular distribution of porous fibers,a pleated fiber filter media model was established based on Voronoi-Random algorithm,and its solid volume fraction(SVF)was changed by encrypting or reducing the number of fibers.Two collision models"Caught on first touch"and"Hamker"were used to simulate the gas-solid two-phase flow of pleated fiber filter media.The results show that the numerical values of pressure loss and filtration efficiency were in good agreement with those calculated by empirical correlation,and the error was within 15%;Comparing the filtration efficiency obtained by two collision models of"Caught on first touch"and"Hamker"with the calculated value of empirical correlation,it was concluded that the filtering efficiency obtained by"Hamker"collision model accords with the reality;The particles deposited on the filter media were not completely captured by fibers,but most of them were captured by dendritic structures formed;The pressure drop and deposition per unit area increased exponentially with the increase of filtration time.