Fractal-Based Model For Evaluating The Filtration Efficiency Of The Non-Woven Fibrous Composites

The classic single-fiber efficiency model based on ordered packing of fiber composites has been widely used to predict filtration efficiency of fibrous filters for 40 years. However, the simplified single-fiber model often overestimates the filtration efficiency as most fibrous composite filters are composed of randomly distributed fibers. The numerical methods have been successfully employed to re-construct the realistic fiber composites, but simulation of the multi-mechanism filtration process in the complex fibrous architecture is computationally expensive and case-based and derivation of a compact versatile model for broad applications remain challenging. In this work, a fractal-based homogenization model is developed to predict the filtration efficiency of the fibrous composites, by considering the spatially randomly distributed fibers and the quasi-random distribution of pore size. A comparison with available experimental results collected from references shows that the proposed model is of high accuracy in predicting the filtration of submicron aerosol particles. The calculated results show that the Kuwabara hydrodynamic factor is sensitive to the pore fractal dimension. The filtration efficiency for different particle diameters can be divided into three stages, which increase as the proportion of fibers warped by the fluid decreases. In addition, the total efficiency decreases with the increasing pore fractal dimension in a nonlinear trend. For the given fractal dimension, the filtration efficiency increases with the increase of pore size ratio, indicating that the more uniform the pore size distribution, the higher the efficiency.