Computational design of locally resonant acoustic metamaterials

Abstract

The so-called Locally Resonant Acoustic Metamaterials (LRAM) are considered for the design of specifically engineered devices capable of stopping waves from propagating in certain frequency regions (bandgaps), this making them applicable for acoustic insulation purposes. This fact has inspired the design of a new kind of lightweight acoustic insulation panels with the ability to attenuate noise sources in the low frequency range (below 5000 Hz) without requiring thick pieces of very dense materials. A design procedure based on different computational mechanics tools, namely, (1) a multiscale homogenization framework, (2) model order reduction strategies and (3) topological optimization procedures, is proposed. It aims at attenuating sound waves through the panel for a target set of resonance frequencies as well as maximizing the associated bandgaps. The resulting design's performance is later studied by introducing viscoelastic properties in the coating phase, in order to both analyse their effects on the overall design and account for more realistic behaviour. The study displays the emerging field of Computational Material Design (CMD) as a computational mechanics area with enormous potential for the design of metamaterial-based industrial acoustic parts.

Multiresonant Layered Acoustic Metamaterial (MLAM) solution for broadband low-frequency noise attenuation through double-peak sound transmission loss response

“…As expected, the more conventional acoustic metamaterial configuration, i.e. consisting of a single resonator/unit cell design, produces an STL response characterized by an attenuation peak followed by a transmission dip on frequencies that, for plane waves at normal incidence, coincide with the associated bandgap limits [18,19] (see Fig. 5(a)).…”

Broadband and intense sound transmission loss by a coupled-resonance acoustic metamaterial

“…Several proposals have been presented to widen the band gaps of locally resonant acoustic/elastic metamaterials (see, e.g., Refs. [22][23][24][25]); however, all these strategies are constrained by the fundamental limitation of a single, narrow attenuation peak in the band gap as directly observed in the imaginary wave vector portion of the dispersion band structure [26].…”

Computational design of innovative mechanical metafilters via adaptive surrogate-based optimization

“…Leaving aside the exceptional possibility to achieve explicit analytical (invertible) solutions of the direct problem -for instance recurring to asymptotic approximations [16][17][18] -numerical strategies based on machine learning can be suitably adopted to solve the optimization problem of maximizing certain spectral properties of interest (inverse problem). Considering the pressing technological demand of low-cost mechanical filters for low and extralow frequency vibrations and noises, a highly-desirable target can be identified in opening spectral stop bands with the largest possible amplitude and the lowest possible centerfrequency [15,[19][20][21][22][23].…”