Large-scale computational design and simulation of viscoelastic metastructures for vibration attenuation

Low-frequency vibration generated by rotating machinery in submarines pose a persistent challenge for acoustic stealth, as they are readily detected by passive sonar and are poorly mitigated by conventional passive or active control technologies under variable operating conditions. This paper introduces a scalable computational framework that couples shape-grammar-driven generative design with finite element simulations to systematically explore viscoelastic metastructures for vibration attenuation. The generative design method is defined by six dimension-independent geometric parameters, enabling automated synthesis of a broad and non-intuitive design space beyond traditional unit-cell parameterizations. Vibration transmissibility of each metastructure is quantified through numerical simulations, which are validated against experimental measurements on representative specimens. The results reveal viscoelastic metastructures exhibiting pronounced and tunable low-frequency attenuation bandwidths, therefore providing enhanced attenuation compared to conventional designs. The resulting dataset establishes a structured mapping between geometry and dynamic response, offering new insight into geometry-driven vibration mitigation mechanisms. Beyond forward analysis, the proposed framework provides a scalable foundation for data-driven and inverse design of metastructures targeting robust low-frequency vibration attenuation. • Scalable generative design of sophisticated metastructures for vibration attenuation. • Automated design-simulation enables large-scale metastructure discovery. • Data-driven links between geometry and damping reveal high-performance designs.