Time-resolved high-dynamic-range particle image velocimetry using local uncertainty estimation

Abstract

The dynamic velocity range of particle image velocimetry is the ratio of the maximum to the minimum resolvable displacement. Although many techniques have been developed in recent years to extend the dynamic range, flows with a wide velocity range still challenge conventional particle image velocimetry. Using multiple-pulse-separation acquisition and a new criterion for the local optimal pulse separation, this paper presents a novel time-resolved high-dynamic-range particle-image-velocimetry methodology. The algorithm maximizes a vector quality metric combining the correlation peak ratio with the estimated local displacement uncertainty and magnitude, expressed as a modified signal-to-noise ratio. Using an axisymmetric turbulent jet as a benchmark case, significant enhancements are shown in the measured turbulence intensity and signal-to-noise ratio throughout the flowfield, but especially in the entrainment region and the outer shear layer. For this case, high-dynamic-range particle image velocimetry increases the dynamic velocity range by 16.5 times compared to conventional double-frame particle image velocimetry. Hot-wire anemometry is used in characteristic flowfield locations as a reference measurement. The results show that time-resolved high-dynamic-range particle image velocimetry automatically selects the optimal pulse separation in each location as a function of time. The method relies only on readily available data, has a low computational cost, and is fully compatible with conventional multigrid vector evaluation algorithms.