3-D microscale simulation of dust-loading in thin flat-sheet filters: A comparison with 1-D macroscale simulations

Abstract

In this work, a microscale approach is undertaken to simulate the instantaneous pressure drop and collection efficiency of fibrous media exposed to particle loading, i.e., filter aging. The air flow field through 3-D disordered geometries representing the internal microstructure of a fibrous filter is obtained by numerically solving Stokes' equations. A Lagrangian approach is used to track the trajectory of particles through our virtual filter media and determine the filter's collection efficiency under different dust-load conditions. The calculations were conducted using the ANSYS CFD code enhanced with a series of in-house C++ subroutines. To better illustrate the value of such CPU-intensive 3-D microscale modeling, we compared the results of our simulations with those obtained from a 1-D macroscale model developed based on some of the pioneering studies reported in the literature. It was found that while the 1-D macroscale models can provide fast predictions for the pressure drop and collection efficiency of a given filter, they require a series of empirical correction factors or case-specific assumptions that limit their usage for design and development of new filter media. The 3-D microscale simulation methods, in contrast, are self-sufficient as they are developed based on first principles. With the current rate of progress in developing high-speed computers, it is expected that 3-D microscale simulations will be the preferred method of filter design in the near future.