Comparison of experiments on bio-inspired hover kinematics with the unsteady vortex model and CFD

In this paper we compare the force and Particle Image Velocimetry (PIV) measurements of biologically inspired hover kinematics to results of an unsteady aerodynamic vortex model (UAVM) and a Navier-Stokes (NS) solver. The baseline Reynolds number and the reduced frequency are 4:8 10 and 0:38, respectively. We consider three versions of the hovering kinematics measured from an Agrius Convolvuli: i) without elevation angle, ii) elevation angle accounted in the pitch angle; and iii) pure sinusoidal pitch-plunge neglecting higher harmonics. The NS computations show good qualitative agreement with experiments with consistent underprediction. The time-averaged thrust coe cient obtained using NS computations are 15% to 18% of the corresponding force measurements. The standard deviation of time history of thrust coe cients, also normalized by the measured time-averaged values, is 13% to 20%. The underprediction is possibly due to blockage e ects in the experiments or excessive dissipation, also re ected in lower values of the vorticity compared to the PIV measurements. The UAVM captures some of the peaks in a qualitative manner. The relative di erence in the time-averaged forces and standard deviation are 8% to 18% and 66% to 93%, respectively. The di erences in prediction of time histories are not re ected in the estimation of time-averaged forces due to cancellation e ects, wherein the forces are underpredicted in the rst half of the stroke and overpredicted in the second half. The discrepancies are attributed to the simplifying assumptions in the UAVM due to which the vorticity in the leading-edge vortex is overpredicted and signi cant di erences in the wing-wake interaction are also noted.