Modular Multi-Scale Digital Twin Framework for Sustainable Design of Hybrid Air-Purifying Facepiece

This study presents a multiscale digital twin framework optimized for the modular, customized construction of a hybrid air-purifying facepiece. Our approach establishes a modular digital library by independently modeling individual components, including the reusable knitted facepiece and biodegradable filter media made from poly(lactic acid) (PLA) and mulberry cellulose. The proposed dual-modeling strategy achieves high prediction fidelity, where resolved modeling applies for the macro-scale knit (150–600 μm) and parametric modeling applies for the microscale nonwoven networks (5–20 μm). Mean relative errors (MRE exp/sim ) in prediction were below 11% across morphological parameters, airflow resistance, and penetration. Crucially, this modularity addresses the intricate interlayer flow dynamics in hybrid facepiece architectures, yielding accurate simulations of any customized layer assembly. This system serves as a robust predictive tool, facilitating the dynamic optimization of stacking strategies (1L to 5L) tailored to specific pollutant profiles (PM 1, PM 2.5 ) without the need for exhaustive physical prototyping. Furthermore, integrating the DOZN sustainability analysis reveals that while filtration efficiencies are comparable, the electrospun PLA web offers superior environmental sustainability over the mulberry-based webs in terms of resource intensity. By virtualizing individual layers into a modular platform, this study provides a comprehensive framework for designing high-performance, eco-conscious air-filtering facepieces tailored to diverse environmental conditions.