Time-Resolved Two- and Three-Dimensional Measurements of Transitional Separation Bubbles

Abstract

Abstract. Advanced methods of the Particle-Image Velocimetry (PIV) technique such as Time-Resolved and Stereo-Scanning PIV have been applied to investigate the separation bubbles on top of a finite circular cylinder and on the suction side of an SD7003 airfoil. A kind of three-dimensional vortex-shedding mechanism associated with the inviscid Kelvin-Helmholtz-instability has been identified to govern the flow field in both cases. The time-averaged velocity distribution is strongly affected by these vortex cascades although the unsteadiness of the bubbles due to the band of frequencies in the spectra of the vortex-shedding mechanism is hidden. Several analyzing methods are used and/or developed and the capabilities of more sophisticated experimental investigations of the three-dimensional flow field associated with separation bubbles are outlined.

Numerical investigation of the interaction between laminar to turbulent transition and the wake of an airfoil

“…Separation (x s ) and reattachment (x r ) points are determined with the null skin friction coefficient, while the length of the LSB (l s ) is defined as the distance between those two points and its height (h s ) as the maximum height obtained by the δ 99 criterion. The present computation agrees well with experiments of Burgmann and Schröder [35]. Though the separation point is slightly advanced, the LSB length and height are within the range of the experimental values.…”

“…While the previous validation considered an infinite twodimensional domain, the investigation of the transition at Re = 20 000 is performed using a three-dimensional confined domain. Hence, in order to identify the effect of confinement, a comparison between the flow at Re = 10 000 in the two-dimensional infinite domain and the one in the three-dimensional confined domain, similar to the experimental one used by Burgmann and Schröder [35], is presented. Fig.…”

“…Computations have been performed in a three-dimensional confined domain (similar to the experimental one of Burgmann and Schröder [35]) for Reynolds numbers ranging from Re = 5000 (laminar) up to Re = 20 000 (transitional). The effect of an increase of the Reynolds number is shown in Fig.…”

“…14(b). Table 2 provides the characteristics of the LSB along with comparisons against two sets of measurements from Burgmann et al [18] and Burgmann and Schröder [35]. Separation (x s ) and reattachment (x r ) points are determined with the null skin friction coefficient, while the length of the LSB (l s ) is defined as the distance between those two points and its height (h s ) as the maximum height obtained by the δ 99 criterion.…”

“…As long as the reattachment point is concerned, full reattachment is observed at the trailing edge (x/c = 1), which was also found by the experiments of Olson et al [20] and recently by Nati et al [21] for a higher Reynolds number (Re = 30 000). Olson et al [20] have also compared their results with Burgmann and Schröder [35] and Burgmann et al [18]. They explain the discrepancies in the LSB characterization by uncertainties due to determination of the angle of attack, to uncertainties due to near-wall measurement resolution, as well as the sensitivity to the Reynolds number and the effects of free-stream turbulence.…”