The aim of this thesis is the characterization and modeling of the longitudinal structures actuated by vortex generators. Results from generic studies performed at low Reynolds numbers have shown that the device induced vortices possess helical structure of the vortex core. Further, their ability to control separation and downstream evolution across the chord of a circular sector have been studied. Similar flow structures to the ones found in the generic experiments have been found in a higher Reynolds number setting, more applicable to realistic cases common to, e.g., aeronautical applications. The helical structure of the vortices can, however, not be confirmed by the results of these experiments due to practical concerns of obtaining a measuring signal with high enough quality and resolution.
Furthermore, in order to study the dynamics of the device induced structures, power spectra from LDA time series have been constructed from the burst-mode LDA theory developed mainly by Buchhave and George [19, 46]. In the process of applying this theory to the LDA time series, a technique has been developed correcting for the effect of random noise in spectra and correlations. The power spectra obtained from the flow behind the actuating devices did not display any distinct periodicity of the flow, but rather a random, or at best quasi-periodic, behavior. In addition, commonly employed interpolation and resampling methods for estimating power spectra from LDA data were compared to the corresponding spectra derived from hot-wire data. When the flow was well resolved, these methods showed acceptable results at high LDA data rates at all frequencies except at the highest ones. However, they failed miserably at low data rates, essentially burying the entire spectrum in frequency dependent noise beyond recognition.
|Place of Publication||Kgs. Lyngby, Denmark|
|Publisher||Technical University of Denmark|
|Number of pages||232|
|Publication status||Published - Nov 2009|
|Series||DCAMM Special Report|