A Preliminary Spectral Decomposition and Scale Separation Analysis of a High-Fidelity Dynamic Stall Dataset

摘要

Dynamic stall results in major excursions from the desired aerodynamic performance. Among the key events are flow separation and the formation of a leading edge vortex (LEV). In this work, we use data obtained from anexten-sively validated Large Eddy Simulation of a plunging SD 7003 airfoil to explore the spectral content and predominant length scales in the LEV, as well as the dynamically significant trailing edge vortex (TEV). The vortex detection is accomplished with an established approach which is shown to perform quite accurately for this highly turbulent flow. Two types of decomposition, Dynamic Mode Decomposition (DMD) and Empirical Mode Decomposition (EMD), are employed to analyze the spatio-temporal scales. Reconstruction is used as a measure of accuracy to identify scales with the highest contributions. A new approach combining EMD with FFT is examined to perform scale separation and spectral decomposition of the flow.