Acoustic Characteristics of Microparticle- Modified Cementitious Composites with Embedded Tesla Valve-Inspired Metamaterials

Abstract

Achieving efficient low-frequency sound absorption in concrete without compromising compressive strength remains a critical challenge in construction materials design. This dissertation addresses this challenge through the development of multifunctional cementitious composites that integrate engineered pore architectures and industrial byproducts to achieve a balance between acoustic performance, structural capacity, and environmental sustainability. To overcome the excessive material thickness typically required for low-frequency sound absorption, a hybrid acoustic strategy was developed using surface-perforated mortar embedded with Tesla valve–inspired acoustic metamaterials and Helmholtz-type resonant cavities. These geometries, fabricated via additive manufacturing, significantly modified internal acoustic pathways, resulting in a 64% reduction in sound reflection and a 68.75% increase in broadband sound absorption, particularly within the 250–2500 Hz frequency range. Sustainability was further enhanced through the incorporation of supplementary cementitious materials (SCMs), including hollow glass microspheres (HGMs), cenospheres (CS), and rubber powder (RP), as partial replacements for cement or sand. These materials reduced environmental impact while influencing hydration behavior and acoustic response. HGM3 accelerated early-age reactions, CS improved internal curing and thermal stability, and RP, despite inducing a 20–25% reduction in compressive strength, proved effective for non- structural sound-absorbing applications. Durability performance was assessed through thermal resistance, chloride penetration, carbonation, and electrical conductivity tests, demonstrating that the optimized composites satisfy relevant standards for lightweight structural materials (ACI, ASTM C330). To further refine and generalize acoustic performance, a data-driven optimization framework combining COMSOL-based thermoviscous simulations and machine learning models was implemented. This framework accurately predicted sound absorption behavior and identified optimal geometric configurations beyond the experimental design space. Overall, this dissertation presents a scalable and sustainable cementitious material platform that integrates acoustic metamaterial design, durability, and data-driven optimization, advancing the development of next-generation sound-absorbing concrete for noise-sensitive infrastructure.