Experimentally informed computational modeling of filtration efficiency of surgical mask and N95 respirator under standard testing and human breathing conditions

Accurate prediction of mask filtration efficiency (FE) is challenging because particle capture depends on the coupled effects of filter structure, particle size, flow rate, electrostatic charge, and humidity. Standard testing provides useful benchmarks but may not fully represent human breathing conditions. This study developed an experimentally informed filtration model to predict the FE of a multilayer surgical mask and an N95 respirator under TSI 8130 A standard testing and representative breathing conditions. The model incorporated measured pressure drops, layer-specific filter properties, and single-fiber capture mechanisms, including diffusion, interception, impaction, and electrostatic attraction. It reproduced measured pressure drops and FEs for both masks and correctly captured canonical filtration behavior, including the most penetrating particle size (MPPS), high flow-dependence of FE, and the dominant role of electrostatic capture. The model also revealed unconventional behavior under cough-like conditions at 360 L/min, including non-monotonic FE–flow-rate dependence for the surgical mask, an MPPS shift from 0.3 µm to 0.2 µm, and near-complete capture of particles ≥0.5 µm by both filters. These results suggest that the National Institute for Occupational Safety and Health-certified FE at 85 L/min provides a conservative estimate of mask filtration performance. The model provides a platform for interpreting experiments and optimizing multilayer mask design under realistic respiratory conditions.