Title: Nonreciprocal model swimmer at intermediate Reynolds numbers

Abstract: Metachronal swimming, the sequential beating of limbs with a small phase lag, is observed in many organisms at various scales, but has been studied mostly in the limits of high or low Reynolds numbers. Motivated by the swimming of brine shrimp, a mesoscale organism that operates at intermediate Reynolds numbers, we computationally studied a simple nonreciprocal 2D model that performs metachronal swimming. Our swimmer is composed of two pairs of paddles beating with a phase difference that are symmetrically attached to the sides of a flat body. We numerically solved the Navier-Stokes equations and used the immersed boundary method to model the interactions between the fluid and swimmer. To investigate the effect of inertia and geometry, we performed simulations varying the paddle spacing and the Reynolds numbers in the range $Re = 0.05 - 100$. In all cases, we observed back-and-forth motion during the cycle and a finite cycle-averaged swim speed at steady state. Interestingly, we found that the swim speed of the swimmer has nonmonotonic dependence on $Re$, with a maximum around $Re\approx1$, a flat minimum between $Re=20-30$ and an eventual increase for $Re>35$. To get more insight into the mechanism behind this relationship, we first decomposed the swim stroke of the swimmer and each paddle into power and recovery strokes and characterized the forward and backward motion. Unlike for reciprocal swimmers, here, for parts of the cycle there is competition between the two strokes of the paddles - the balance of which leads to net motion. We then studied the cycle-averaged, as well as, power-stroke-averaged and recovery-stroke-averaged fluid flows, and related differences in the fluid field to the nonmonotonic behavior of the swim speed. Our results suggest the existence of distinct motility mechanisms that develop as inertia increases within the intermediate-Re range.