Adjustable fiber structure design for precise aerosol filtration

High-efficiency particulate air (HEPA) filters are widely used to filter aerosols. Predicting a HEPA filter's performance is essential to guarantee occupant health and comfort. Traditional cell models do not consider the fiber diameter distribution and spatial orientation, which makes them unable to predict filter performance well. We use the computational fluid dynamics (CFD) method to realize the initial performance simulation of glass-fiber (GF) filter media based on a random 3-D fiber model. In this study, we divided the filter media into multiple independent zones in the thickness direction and generated a fiber structure with the same fiber diameter distribution as the real filter media. Based on this method, we investigated the stability of the model and the effects of parameters such as computational domain width, fiber windward angle, filtration velocity, and fiber diameter distribution on the filtration performance of the filter media. The simulated pressure drop is in good agreement with the experimental values, and the efficiency of the most penetrating particle size (MPPS) of the filter media can be accurately calculated. A fiber model design and performance comparison of different fiber diameter distributions are realized. We believe this method can help us better understand the particle capture process and design a fiber structure for filter media to obtain better performance. [Display omitted] •Established a random 3-D fiber model with adjustable parameters.•Fibers' windward angle has little effect on filtration performance.•It's possible to increase quality factor through fiber diameter distribution design.