ETD system

Electronic theses and dissertations repository

 

Tesi etd-07192007-154729


Thesis type
Tesi di dottorato di ricerca
Author
Iungo, Giacomo Valerio
email address
giacomo.iungo@ing.unipi.it, iungogv@hotmail.com
URN
etd-07192007-154729
Title
Investigation on the dynamics of wake vorticity structures and their experimental evaluation
Settore scientifico disciplinare
ING-IND/06
Corso di studi
INGEGNERIA AEROSPAZIALE
Commissione
Relatore Prof. Buresti, Guido
Relatore Prof. Lombardi, Giovanni
Parole chiave
  • triangular prism
  • tip vortex
  • three component hot wire anemometer
  • rapid scanning
  • five hole probe
  • bluff body aerodynamics
  • wandering
  • wavelet transform
Data inizio appello
20/07/2007;
Consultabilità
completa
Riassunto analitico
The present work gives a contribution to the investigation on the dynamics of wake vorticity structures and to the development of measuring techniques and data correction methods in order to perform their proper experimental evaluation. The topic of this Ph.D. thesis consists of two main projects: survey of the wake generated from a triangular prism of moderate aspect ratio in cross-flow and investigation into the effects of wandering on static velocity measurements of a wing-tip vortex. As for the first subject, a deeper understanding of the physical mechanisms giving rise to fluctuations of the flow field connected with the wake vorticity structures is reached. The velocity fluctuations detected in the wake of a prism having triangular cross-section and an aspect ratio H/w=3.0 are characterized. Preliminarily, flow visualizations with injected smoke and a laser sheet were performed to obtain clues about wake morphology and dynamics. Single sensor hot-wire anemometer measurements were carried out in the flow field as well as pressure measurements on the free-end and on the rear surface of the model. It is shown that, besides the fluctuations induced by an alternate vortex shedding with Strouhal number, St¡Ö0.16, further components are present, with different relative intensities in different wake regions. A spectral contribution at St¡Ö0.05 is found to dominate all the velocity signals in the upper-wake and it is attributed to a vertical, in-phase oscillation of a couple of counter-rotating axial vortices detaching from the front edges of the model free-end. An intermediate component is also found, occurring at St¡Ö0.09; the analysis of a previously available LES simulation was fundamental for the interpretation of the physical mechanism giving rise to this flow fluctuation, which is associated with the oscillations of a transversal shear layer detaching from the rear edge of the model free-end. Proceeding downstream, it bends downwards into the wake in such a way to be reversed upstream impinging the rear surface of the model. Consequently, a recirculation region is delimited by this transversal shear layer. This feature is also assessed from the pressure measurements carried out on the model surfaces; indeed, a pressure maximum is ascertained on the rear surface at z/H=1/3 and fluctuations at St¡Ö 0.09 are singled out just at the locations below the recirculation region. Furthermore, the statistics of this frequency are comparable to the ones related to the same spectral component singled out in proximity to rear edge of the free-end, and thus most probably the fluctuations observed in the two zones are generated from the same vorticity structure, viz. the transversal shear layer. From the numerical visualizations of the vorticity field it is observed that the fluctuations of the recirculation region are strictly connected with the vortex shedding. Lateral vorticity sheets are dragged in the upper wake generating in correspondence to the wake symmetry plane a direct vertical \action&#34; on the transversal shear layer. Most probably this intricate wake morphology is the physical mechanism giving rise to the oscillations of the recirculation region. Furthermore, it is experimentally assessed that modifications on the vertical edges of the model generate a variation of the vortex shedding frequency comparable to the one produced on the fluctuation frequency of the transversal shear layer. However, no variations were found in the fluctuations at the lower frequency in the upper part of the wake, which suggests that they are likely to be essentially connected with an instability of the axial vortices originating from the free-end.<br>The second project is an investigation on the effects of wandering on static measurements of a wing-tip vortex and on the correction of the measured velocity fields. Wandering consists in random oscillations of the vortex core; consequently, vortices evaluated by static measuring techniques appear as more diffuse and weaker than in reality. Numerical simulations of the wandering of a Lamb-Oseen vortex were performed by representing the vortex core locations through a bi-variate probability density function. It was found that wandering amplitudes smaller than 60% of the core radius are well predicted by using the ratio between the RMS value of the azimuthal velocity and its slope measured at the mean vortex centre. With increasing wandering amplitudes the predictions become more inaccurate, showing errors of 35% of the actual value for wandering amplitudes comparable to the core radius. Furthermore, from the numerical simulations it was found that the principal axes of wandering are well predicted from the opposite of the cross-correlation coefficient between the spanwise and the normal velocities measured at the mean vortex centre. Four different algorithms were then applied to correct mean velocity fields for wandering smoothing effects. All the methods perform the deconvolution of the mean velocity field with the bi-variate probability density function, that represents the wandering. The methods have the advantage of avoiding any assumption or any fitting of the mean velocity field. The performed corrections were very accurate for the simulations with wandering amplitudes smaller than 60% of the core radius, whereas errors become larger with increasing wandering amplitudes (up to an error of 15% of the actual value on the correction of the peak azimuthal velocity for wandering amplitudes comparable to the core radius). Subsequently, the whole procedure to evaluate wandering from static measurements and to correct the mean velocity field for wandering effects was applied to the data of a tip vortex generated from a NACA 0012 half-wing model; these data were obtained using a five hole probe and a three component hot film anemometer. The static measurements corrected for wandering effects were then compared to measurements carried out through rapid scanning. This technique consists in traversing the five hole probe, fixed on a rotating arm, through the vortex core with a sufficiently high velocity in order to consider the vortex as roughly fixed during each scan. Consequently, these measurements are theoretically not affected by wandering. Tests were performed to investigate on the behaviour of the wandering by varying the streamwise distance, the wing angle of attack or the Reynolds number. Firstly, wandering was found to be not a self-induced phenomenon; indeed, its amplitude was reduced with increasing vortex strength. The latter seems to be the principal vortex parameter to control the wandering, as neither the downstream distance, the wing angle of attack or the free-stream velocity have an absolute influence on the wandering. In other words, the wandering amplitude can be reduced by increasing the wing angle of attack, the free-stream velocity, or reducing the streamwise distance from the wing, but if the vortex is sufficiently strong it may be completely insensitive to the variation of these parameters. All the tests were performed with the same turbulence level of the free-stream, and thus it might be more suitable affirming that the principal parameter controlling the wandering could be the ratio between the strength of the vortex and the free-stream turbulence level.<br>
File