ABSTRACT

The concept of metamaterials has emerged as an exciting frontier. These are engineered structures designed to possess unique properties, often unattainable in nature. This concept has not only captured the imagination of the scientific community but has also become a focal point of interest within various industries. Among the diverse applications of metamaterials, the field of acoustics stands out prominently. The ability to create materials with precise and efficient noise attenuation in targeted frequency ranges holds profound significance across a spectrum of different sectors.

Within this landscape of metamaterial exploration, multiresonant layered acoustic metamaterials (MLAM) have surfaced as a particularly promising solution. Distinct from conventional locally resonating acoustic metamaterials, MLAMs leverage the intricate coupling of resonances within layered structures, showcasing exceptional sound-blocking capabilities across broader frequency ranges. Through its design, MLAM panels successfully address two pivotal challenges in acoustics. Firstly, they excel in attenuating sound within specific broadband frequency ranges, surpassing the efficiency of conventional solutions. Secondly, the layered structure of MLAMs not only enhances attenuation through novel coupling mechanisms but also facilitates seamless and cost-effective large-scale production. This inherent design feature positions MLAMs as a viable and scalable solution, overcoming the production limitations encountered by other acoustic metamaterials.

Recognizing the pivotal role of structural design in shaping the properties of acoustic metamaterials, this doctoral thesis presents significant advancements in the design and optimization of these materials, with a particular focus on MLAM and their innovative applications. Key contributions of this research include the development of advanced design paradigms, the integration of machine learning in design optimization, and innovative structural designs.

The research introduces a pioneering computational framework based on multiscale homogenization, enabling rapid and precise evaluation of sound transmission loss (STL) across various frequency spectra. This framework facilitates the parameterization and optimization of MLAM designs, accommodating practical constraints such as weight, thickness, and geometric tolerances. By incorporating machine learning techniques, this work enhances the modeling and optimization processes of acoustic metamaterials. Machine learning algorithms are employed to streamline the design process, significantly reducing the time and computational resources required to achieve optimal acoustic properties. Inspired by MLAM, new configurations have been proposed with the objectives of enhancing attenuation ranges and levels (namely n-MLAM and MLAM+), and of developing much lighter and thinner solutions (TCAM). The n-MLAM and MLAM+ explore different multi-coupling resonance mechanisms, broadening the effective sound attenuation bandwidth and enhancing sound insulation properties. The TCAM utilizes a different coupling mechanism through a bending mode, which enables the development of much lighter and thinner designs.

In summary, this thesis represents a comprehensive investigation into the advancements in acoustic metamaterial design and optimization. The research findings contribute to the development of efficient, scalable, and practical solutions for sound attenuation, addressing critical challenges in various industrial applications.

Committee

• President: Dr. Oriol Guasch Fortuny
• Secretary: Dr. Teresa Pàmies Gómez
• Member: Dr. Francisco Javier Montans Leal
  
  
• PhD advisor: Prof. Xavier Oliver

PHD CANDIDATE

Mr. Gastón Sal is a civil engineer with the Mechanics of Advanced Materials and Metamaterials research cluster at CIMNE.