Conditional large-scale structure of a high Reynolds number turbulent boundary layer

A spanwise array of 10 surface mounted hot-film shear-stress sensors coupled with a traversing hot-wire probe are used to identify the conditional structure associated with a large-scale skin-friction event in a high Reynolds number turbulent boundary layer (Reτ 14000). Instantaneous shear-stress data indicate the presence of large-scale structures at the wall that are comparable in scale and arrangement to the superstructure events as reported previously by Refs (4; 5). Conditional averages of streamwise velocity computed based on a low skinfriction footprint at the wall offer a wider three-dimensional view of the average superstructure event. These events consist of highly elongated forward-leaning low-speed structures, flanked on either side by high speed events of similar general form. An analysis of small-scale energy associated with these large-scale events reveals that the small-scale velocity fluctuations are modulated by the presence of large-scale features. In general it is observed that the attenuation and amplification of the small-scale energy seems to approximately align with largescale regions of streamwise acceleration and deceleration respectively. These results suggest that small-scale structures, including near-wall streaks/vortices, are influenced by the passage of outer layer large-scale events. Therefore, any control strategy for high Reynolds number wall-bounded turbulence that aims to control the small-scale activity in the near-wall region will likely need to account for the behaviour of the large-scale structures that are present in the outer layer. The results might also suggest the viability of specifically targeting the large-scale structure in order to control turbulence. Facility & Measurement Array Experiments are performed in the High Reynolds Number Boundary Layer Wind-Tunnel (HRNBLWT) at the University of Melbourne, an open-return blower wind-tunnel with a working section, 2 1 27 m. Full details of the facility are available in ref (10). Measurements were performed in the turbulent boundary layer developing over the tunnel floor approximately 21 m downstream of the tripped inlet to the working section. Freestream velocity (U∞) was 20 33 ms 1 (and freestream turbulence intensity u∞ U∞ 0 2%). The boundary layer thickness (δ at the measurement location was 0.326 m yielding a Kárman number, Reτ δUτ ν 14200 (where, ν is the kinematic viscosity and Uτ is friction veloicty). Throughout this paper, x, y and z will be used to denote the streamwise, spanwise and wall-normal axes, with u, v and w denoting the respective fluctuating velocity components. Capitalised velocities (e.g. U) or overbars (e.g. u 0) indicate time-averaged values. Angle brackets (e.g u ) denote conditionally averaged quantities. The superscript is used to denote viscous scaling of length (e.g. z zUτ ν), velocity (U U Uτ) and time (t tU2 τ ν). A spanwise array of 10 flush-mounted hot-film sensors are affixed to the tunnel wall 21m downstream of the inlet to the CL traversing probes y x z spanwise array ∆yhf 0 08δ 10 9 8 7 6 5 4 3 2 1 FLOW Figure 1: Diagram detailing the measurement array. working section. This array covers a spanwise domain of 0 7δ, with a spanwise resolution ∆yhf 0 026 m or 0 08δ. The 10 Dantec 55R47 glue-on flush-mounted sensors are operated in constant temperature mode using AA labs AN1003 anemometers with overheat ratio (OHR) set to approximately 1.05. The active spanwise length of the sensor lhf (0.9 mm) equates to a viscous-scaled length l hf 39 for this experiment. The sensors are numbered sequentially from 1 to 10 as indicated in figure 1. Two hot-wire probes are mounted above the spanwise array in a wall-normal traverse. Two sensors were employed to minimise inaccuracy due to calibration drift, but for this report only the data from the left hand-side sensor in figure 1 is analysed (this probe is located almost exactly above hot-film sensor 6). This hot-wire probe has an etched sensor length of l 0 5 mm (equating to l hw 22 for the current experimental conditions). The suffixes ‘hf’ and ‘hw’ will be used to denote the hot-film and hot-wire sensors, respectively. Instantaneous fluctuations from the spanwise array Hot-film sensors are known to suffer from reduced frequency response (as compared to hot-wires) due to heat conduction to the substrate (see 3). A careful comparison between the hotfilm measured spectra and that measured by hot-wire sensors located very close to the wall, reveals that for frequency content 50Hz there is an increasing attenuation of the hot-film measured statistics. Assuming a constant convection velocity (discussed below), this equates to attenuation of streamwise lengthscales 0 5δ. In light of this analysis, the signals from the hotfilm sensors are filtered using a 2-D gaussian (with approximate length 0 5δ and spanwise width 0 16δ) to leave only the largescale component of the signal. As a precursor to these experi−1 0 1 PSfrag replacements