High-efficiency filtration: Smart designs for particle trapping

Abstract

Trapping solid particles within fluid flows is a critical concern for maintaining the health of living organisms, enhancing the efficiency of industrial equipment, and more. In this study, we present an optimal design for achieving the highest possible particle-trapping rate in a two-dimensional filter with turbulent fluid flow. To achieve this, we use a genetic algorithm to determine the optimal arrangement of square obstacles within a turbulent flow field. The process starts with an image processing method (IPM) that identifies geometric objects in the filtered image. Following this, the edges of these objects are delineated, and a mesh is generated throughout the fluid flow field and around the identified filter objects. To solve the hydrodynamics and turbulent equations, we apply the finite volume method. Furthermore, a staggered grid is utilized to store scalar and vector variables. The genetic algorithm (GA) iteratively generates new arrangements, which are then evaluated and selected for mutation to refine the optimized design. This refined configuration of the filters is designed to enhance the particle trapping rate. A comparison between the optimized filter locations and those of a simpler design reveals a significant reduction in the escape of particles, with a 23 % decrease observed in the optimized condition.