Bio-inspired locomotion of a rotating cylinder pair

We developed bio-inspired reduced-order models of swimmers, consisting of a self-propelling pair of rotating cylinders. The aim of the project is twofold. First, simplified and non-deforming geometries can more easily be employed in small-scale robotic applications to solve relevant engineering problems. Second, they can serve as reduced physics models to efficiently simulate fluid-mediated interactions in schools of swimmers and perform learning studies involving multiple swimmers. In the first half of the thesis, we investigate the self-propulsion regimes of a pair of counter-rotating cylinders. For low rotation rates, the cylinders behave like a vortex dipole and the flow is characterized by an elliptical closed streamline surrounding the cylinders. For intermediate rotation rates, the cylinders move in the opposite direction and each has a different set of closed streamlines. Further increasing the rotation rate, the motion of the pair becomes unstable. We systematically explore the phase space defined by the non-dimensional centre-to-centre distance and the rotational Reynolds number, and find inverted exponential correlations that describe the transition between states. In the second half, we design three different locomotory modes of the cylinder pair, with few degrees of freedom, inspired by the movement of undulatory fish and jellyfish. The parameter spaces were explored with the CMA-ES stochastic optimization algorithm in order to find the best solutions in terms of maximum speed and efficiency. The undulatory fish-inspired motion achieves propulsion by shedding vorticity in a sequence of alternating sign vortices, similarly to its biologic counterpart. The jellyfish-inspired motion during each period sheds a vortex dipole, which generates a strong momentum flux in its wake and thus a large thrust, the definition of the swimming mode of jet-propelled oblate medusae.Solid and Fluid MechanicsProcess and EnergyMechanical, Maritime and Materials Engineerin