Turbulence Interaction with an Airfoil
When a lifting airfoil moves through a turbulent stream (shown above) an unsteady pressure field develops around the airfoil which produces acoustic waves that manifest into broadband noise. Such phenomena can readily be identified in nearly all types of vehicles (e.g. automobiles, planes, helicopters, ships, submarines, etc.) the results of which can be quite troublesome. For example helicopter noise levels can easily approach 100dB with a large portion of this noise related to the rotor blades cutting through the wake of the leading blade. Sophisticated submarine propulsors which are similar in form to an axial compressor in a turbofan engine can produce considerable noise as the propulsor blades slice through boundary layer produced turbulence. In combat scenarios such noise levels obviously hamper the stealth capability of these vehicles. In addition, with the passage of Bills such as the “Silent Skies Act of 1999” industry is increasingly coming under pressure to reduce the levels of noise pollution associated with turbomachinery propulsors (i.e. turbofans, turbojets, etc) on their aircraft. For these reasons sophisticated models addressing the interaction of a lifting airfoil with turbulence are required.
The past 60 years has shown an increase in model complexity from the relatively crude flat plate approximation to the development of more complex numerical, distortion representations considering real airfoil geometry and angle of attack. The goal of any such model is to predict the unsteady pressure field occurring on the surface of the lifting airfoil which in turn allows the estimation of radiated noise levels. This cannot be accomplished without identifying the blade response function which represents the manner in which the airfoil responses to an incoming turbulent flow. With this target in mind researchers set out to develop a realistic and accurate formulation of the problem.
This is study is on the interaction of turbulence with a lifting airfoil. Of particular interest is the unsteady pressure field on the surface of an airfoil immersed in turbulence. Such interaction produces broadband noise and can be found in nearly all modern vehicles. The goal of this research, beyond improved physical understanding of such phenomenon, is to provide a database against which existing broadband noise prediction schemes can be tested and offer insight into to the development of more complex (and therefore realistic) modeling schemes.
To meet these goals several measurements have been made with a NACA 0015 airfoil (pictured to left--1ft chord x 6ft span) which has 96 tiny microphones mounted sub-surface immersed in two different scale turbulent flows. The turbulence is generated with static grids; a large grid, which has cell sizes of 12" produces homogenous turbulence that has an integral scale of 3", a small grid, which has cell sizes of 1" produces near homogenous turbulence with an integral scale of 1/4". These measurements are carried out in the Virginia Tech Stability Wind Tunnel which is the largest university owned wind tunnel facility in the U.S.
We are interested primarily in the effects of angle of attack on the surface pressure field and how the inflow turbulence scale may influence the interaction. Such detailed space-time correlation measurements of the surface pressure field have never been made before.
The results of this measurement are summarized well in the figure to the left. This is a plot of the variation of unsteady lift (the amount of oscillation in lift) with frequency as measured at different angles of attack in both large and small scale turbulence. The large grid results suggest there is a decrease in the amount of unsteadiness in the lift as the angle of attack increases. This is a very unexpected result based on theoretical formulations and previous experimental work. In the small scale turbulence the unsteadiness in lift increases with increasing angle of attack which is the expected result.
We are now in the process of investigating the cause of these two different angle of attack effects. The answer appears to be rooted in the 3-dimensional nature of the interaction which requires the utilization of more sophisticated analysis techniques. A comprehensive study is therefore underway to ensure more sophisticated techniques can be reliably applied to this data set. If this holds true then the data will be decomposed into 2-dimensional wavenumber space where the angle of attack mechanisms hopefully will be revealed.
Data related to this project may be found here - Turbulence Interaction with an Airfoil - DATA.