Bio-inspired acoustic meta-panel for interior noise reduction in rail cabins

This research presents the design, numerical optimization, and experimental validation of two bio-inspired membrane-type acoustic metamaterials—one based on a hexagonal honeycomb topology and the other on a spider-web configuration—for targeted low-frequency sound insulation in railway cabins. Field noise measurements from Fadak passenger trains identified two dominant interior frequency bands: a primary peak near 630 Hz and a secondary band around 1.6 kHz. The proposed meta-panels were modeled as full three-dimensional cellular arrays in COMSOL Multiphysics, with sound transmission loss (STL) computed in 1/6-octave bands and analyzed through modal decomposition. The honeycomb design, characterized by a lower first natural frequency, demonstrated enhanced attenuation at very low frequencies, whereas the spider-web structure exhibited a denser modal distribution, enabling broader suppression near 1.6 kHz. A constant volume/mass parametric analysis was followed by a single-objective Nelder–Mead optimization to maximize STL at 630 Hz by tuning two key geometric parameters. The optimized configurations achieved up to ∼10 dB improvement at the target frequency, with the honeycomb variant offering superior performance-to-mass ratio, and the spider-web variant delivering stronger attenuation peaks at the expense of increased mass. Prototypes fabricated via fused deposition modeling (FDM) using flexible thermoplastic polyurethane (TPU) and thin polyethylene membranes were tested in a coupled reverberant–anechoic setup, confirming the numerical predictions. These findings demonstrate that the integration of bio-inspired geometries with lightweight, flexible membrane structures enables precise control of low-frequency acoustic performance. The proposed designs offer a scalable and weight-efficient solution for mitigating interior noise in railway passenger environments, with potential applicability across a range of transportation and architectural contexts.