Optimal swimming locomotion of snake-like robot in viscous fluids

Abstract

Snake-like robots have several joints that enable them to swim in various types of viscous fluids such as water, mud, and clay by varying the undulation. Because the optimal swimming motion is assumed to be different in fluids with different viscosities, we constructed an equation of motion for a snake-like robot swimming in viscous fluids and attempted to find the optimal swimming motion via numerical analysis. We constructed a snake-like robot comprising eight links in this study that could swim in four fluids with different kinematic viscosities, and measured their joint positions and relative angles during swimming. Accordingly, we proposed a method for identifying multiple unknown parameters of fluid forces acting on a body using the unscented Kalman filter, which could determine the unknown added mass and drag coefficient. Subsequently, we clarified that the inertial drag model is effective in water (small viscosity, μ∼1mPas), while the friction drag model is effective in oil (large viscosity, μ>100mPas). By considering the increase in frictional drag due to boundary layer thinning in the proposed model, we identified the constant drag coefficient regardless of the swimming frequency. Next, we employed the constructed numerical model to explore the trade-off relationship between consumed power and velocity during swimming using exhaustive search. Consequently, we confirmed that the swimming frequency decreased as the fluid viscosity increased under the same swimming power consumption. Moreover, the wavelength of undulation was confirmed to be between half and one wavelength, regardless of the viscosity.