Dynamic stall of rapidly pitching airfoils: MTV experiments and Navier-Stokes simulations

Dynamic stall occurs when an airfoil is rapidly pitched beyond its static stall angle of attack, α_ss. Separation is delayed well beyond α _ss, and an understanding of the boundary layer's evolution in the vicinity of the leading edge is essential to an understanding of the dynamic stall process. For uniformly subsonic flows, the dynamic effects due to pitch rate have been observed to outweigh the effects due to Mach number or Reynolds number, so in this work three different pitch rates are investigated using a low Reynolds number/low Mach number flow. Boundary-layer resolved experimental measurements are presented of the flow field that results when the airfoil's angle of attack increases at a nominally constant rate. These measurements are made using Molecular Tagging Velocimetry (MTV), and an overview of the technique is presented in addition to the experimental results.; Previous numerical simulations of dynamic stall have provided insight into the velocity, vorticity, and pressure fields, but the validation of those simulations against experimental results has typically been restricted to a comparison of the surface pressure distribution or integrated loads. In many cases, validation has also been complicated by a mismatch between the experimental and computational parameters such as Reynolds number or airfoil motion profile. In the present work a fully-compressible, two-dimensional Navier-Stokes solver is used to simulate the dynamic stall flow field using parameters which are essentially the same as those employed in the experiments.; In both the experimental and computational results for all of the pitch rates considered in this study, a detached shear layer first forms near the leading edge when a thin region of positive vorticity develops underneath the negative vorticity of the initial boundary layer. However, this detached condition does not significantly increase the boundary layer's growth rate. A very small second region of negative vorticity subsequently forms between the positive vorticity zone and the wall, after which the boundary layer begins to thicken very rapidly. Although the initial detached shear layer is absolutely unstable, it appears that the higher rate of instability amplification present in the three-layer vorticity field is required to produce the formation, growth, and shedding of the dynamic stall vortex.; Conventional grid resolution studies indicated that the computational results of this study were well resolved, but a comparison of the experimental and computational results suggests that the grid resolution in the streamwise direction must be increased by at least a factor of eight before the computations will actually be able to resolve the length scales which are of importance in this flow field.; Dynamic stall is a specific example of unsteady separation, a topic of extensive study and debate for many decades. The Moore-Rott-Sears criterion is perhaps the most commonly accepted way to determine where and when unsteady separation occurs. Nevertheless, our experimental and computational results both agree with previously published simulations which indicate that MRS points do not exist in the dynamic stall flow field if the frame of reference rotating with the airfoil is the correct one.