Multi-objective optimization of three-dimensional riblet surfaces for hydrodynamic and acoustic performance

Riblets inspired by the dermal denticles of shark skin are widely recognized for their drag-reducing performance. Although previous research has predominantly focused on two-dimensional riblet geometries, three-dimensional (3D) topographies remain underexplored due to the complex architecture of denticle-inspired surfaces. Natural riblet arrays, comprising thousands of interconnected dermal denticles, pose challenges in terms of parameterization, simulation, and fabrication. This work addresses these challenges by introducing a 3D, riblet-reinforced surface topography design that reduces drag, suppresses flow-induced noise, and simplifies both parameterization and prototyping, ultimately providing a scalable solution for towed array sonar applications. Leveraging Bayesian optimization, our computational fluid dynamics (CFD) results reveal that the optimal design decreases the overall sound pressure level by 6.87 dB and reduces drag by 0.34%, effectively balancing noise mitigation with hydrodynamic performance. The design that achieves the greatest noise reduction lowers flow noise by 8.81 dB, albeit with a slight increase in drag. The most effective design for drag reduction yields a 5.18% decrease, accompanied by significant noise suppression across key frequency bands. Flow field analysis demonstrates that our design alters the near-wall vorticity dynamics by promoting the formation of vortex rings that detach from the surface, thereby reducing turbulent energy transfer and limiting sound pressure fluctuations relative to a smooth surface design. To this end, the combination of CFD simulations and Bayesian optimization offers an efficient pathway to refine riblets-reinforced surface topographies, paving the way for advanced bioinspired designs that improve acoustic performance and efficiency in underwater applications.