Modeling of micro-slit /perforated panels in multilayers for efficient design and validation of broadband sound absorbers

This work focuses on design of multilayer micro-perforated/slit panels (MPP/MSP) to create high sound absorption over a wide bandwidth. Theoretically micro-slit panels are equivalent to micro-perforated panels and can similarly achieve high absorption coefficients. Yet, fabrication of microslit panels is much easier. Using transparent panels, they are visually more transparent than MPPs since the MSPs yield more unobstructed panel for a given perforation rate. However, MPP and MSP absorbers in single setting are frequency limited to a bandwidth often not sufficient in applications. Stacking multilayered MPP/MSP panels, a high absorption becomes achievable in a wider bandwidth. Using existing single-panel MPP/MSP models, a transfer matrix method is suitable for cascading multilayers to predict overall absorption performance. This prediction model is applied to a Bayesian framework in coping with design challenges arising from designing MPP/MSP parameters for multiple layers that fulfills a given design scheme. The Bayesian design reported here rapidly converges to a parsimonious number of layers with optimized parameters for each layer, and the overall absorption is able to fulfill the design scheme for minimally possible fabrication complexity. In addition, performing iterative Bayesian analysis during the panel fabrication and experimental validation guides corrections due to fabrication inaccuracies.