High-speed particle image velocimetry for the efficient measurement of turbulence statistics

Abstract

A high frame rate camera and a continuous wave laser are used to capture long particle image sequences exceeding 100,000 consecutive frames at framing frequencies up to 20 kHz. The electronic shutter of the high-speed CMOS camera is reduced to 10 microseconds to prevent excessive particle image streaking. The combination of large image number and high frame rate is possible by limiting the field of view to a narrow strip, primarily to capture temporally resolved profiles of velocity and derived quantities, such as vorticity as well as higher order statistics. Multi-frame PIV processing algorithms are employed to improve the dynamic range of recovered PIV data. The recovered data is temporally well resolved and provides sufficient samples for statistical convergence of the fluctuating velocity components. The measurement technique is demonstrated on a spatially developing turbulent boundary layer inside a small wind tunnel with Re_\delta = 4800, Re_\tau = 240 and Re_\theta = 515. The chosen magnification permits a reliable estimation of the mean velocity profile down to a few wall units and yields statistical information such as the Reynolds stress components and probability density functions. By means of single-line correlation it is further possible to extract the near wall velocity profile in the viscous sub-layer, both time-averaged as well as instantaneous, which permits the estimation the wall shear rate \dot{\gamma} and along with it the shear stress \tau_w and friction velocity u_\tau. This data is then used for the calculation of space-time correlation maps of wall shear stress and velocity.