Hybrid analytical-numerical optimization design methodology of acoustic metamaterials for sound insulation

Acoustic metamaterials are becoming promising solutions for many industry applications, but the gap between theory and practice is still difficult to close. This research proposes an optimization methodology of acoustic metamaterial designs for sound insulation that aims to start bridging this gap. The proposed methodology takes advantage of a hybrid analytical-numerical approach for computing the sound transmission loss of the designs efficiently. As a result, the implementation of optimization techniques on numerical model designs becomes practically possible. This is exemplified with two test cases: (i) optimization of the sound transmission loss of a single gypsum board panel and (ii) optimization of the noise reduction of outdoor HVAC units. Two resonator designs, one used previously for sound radiation in flat panel speakers and the other for enhancing the sound transmission loss at the mass-air-mass resonance of double panels, are here optimized for the two test cases. This shows how an existing resonator can be adapted for new purposes, thus making the design of acoustic metamaterials efficient. The optimized metamaterials outperform the original designs as well as traditional approaches to sound insulation.