Predictive Modeling of Sound Absorption in Porous Materials: Model Complexity, Parameter Identifiability and Multiscale Characterization

Porous materials are widely used for sound absorption, but predicting their acoustic response remains challenging because of the complex interaction between pore morphology and material parameters such as porosity, airflow resistivity, tortuosity, characteristic lengths, density and thickness. This mini-review critically examines the evolution of predictive approaches from empirical formulations, including the Delany–Bazley (DB) and Miki models, to physically based equivalent-fluid models such as Johnson–Champoux–Allard (JCA) and Johnson–Champoux–Allard–Lafarge (JCAL), as well as inverse, numerical, and emerging machine-learning (ML) approaches. Particular attention is given to their applicability to conventional polymeric foams, natural and biomass-based materials, recycled materials and architected porous structures. The review examines the relationship between measurable macroscopic parameters, pore-scale morphology and predicted acoustic response, with emphasis on the limitations associated with material heterogeneity, model assumptions and parameter uncertainty. Inverse acoustic characterization is discussed in terms of parameter non-uniqueness, sensitivity to initial conditions and bounds and the distinction between fitting accuracy and physical identifiability. Overall, increasing model complexity does not necessarily improve predictive reliability. Reliable prediction requires physically meaningful parameters, appropriate model assumptions, independent validation and a balance between model complexity, parameter identifiability, and computational cost.