Instantaneous turbulence profiles in the wake of an oscillating airfoil

Abstract

THE actual spectacular development of computational methods in fluid dynamics has reached a stage where twodimensional, time-averaged Navier-Stokes equations can be solved within reasonable cost and storage requirements, for both steady and unsteady situations. The comparison of calculated Navier-Stokes solutions and experimental data, however, is still limited, mainly by a lack of detailed information about the input to the equations; and the question of the influence of the unsteadiness on the structure is still unanswered. Experimental data in this field were recently compiled by Carr.1 From this review it appears that in thexcase of a pitching airfoil little or no detailed information about velocity and profiles is to be found at the present time, although for the corresponding steady-state case, such information is available but not profuse.2'3 In the present investigation, detailed experimental profiles for the velocity and non-zero Reynolds stresses are presented at different downstream positions in the wake of a sinusoidally pitching NACA 0012 airfoil. Several frequencies, mean incidences, and oscillation amplitudes are considered. Instantaneous incidences vary from zero to just above the static stall limit. The chord Reynolds number is 300,000, the airfoil aspect ratio is 1.6, and a tripping wire is mounted at 10% chord distance from the leading edge. Trailing-edge stall was observed as the static incidence reaches 14 deg. Instantaneous velocity and profiles are available at five downstream positions in the near wake of the airfoil at 48 times during one oscillation period. The data are available from NASA AMES. Contents A single slanted rotating hot wire measurement technique is used for the measurements of velocities and all four non-zero Reynolds stresses. Details of this technique can be found in Ref. 4. Velocities can be considered as accurate within a few percent, except near stall and recirculation areas. The technique is subject to a small turbulence assumption. Turbulence levels can be considered as accurate within 5 % for most experiments, but are less reliable at high levels owing to this assumption. Since the mean flow is twodimensional, two Reynolds stresses can be neglected, and no significant differences were found when neglecting these stresses when processing the data. Mean flow reversal has been detected from the continuity in flow angle profiles. This was made possible by the high density of data within one instantaneous profile. The technique does not allow the observation of random flow recirculations such as vortex shedding.