Experimental Study of Cavitation Inception During Vortex-Vortex Interaction and Investigation of Constrained Volumetric Velocimetry Methods

Abstract

This work is comprised of two topics: a study of artificially constrained volumetric velocimetry and cavitation during the interaction of a pair of counter-rotating vortices. Often, an experimentalist may not have sufficient access to achieve an ideal illumination of a volumetric velocimetry measurement domain. Instead, illumination that is coaxial to the cameras’ lines of sight may be necessary, which is believed to hinder accurate tomographic reconstruction (used in tomographic particle image velocimetry, Tomo-PIV) as it precludes a sharp cutoff in brightness over the depth of field. To probe this difficulty, data is taken in the wake of a triangular wedge (18 mm base height, 15 degree opening angle) placed in a recirculating water channel at 6 m/s. Volumetric velocimetry and planar velocimetry data are taken with a Photonics DM-100 laser (dual cavity laser with 100 mJ/pulse per cavity at 10 kHz) and four Phantom v1212 cameras at a seeding density of 0.03 ppp with sharp cutoffs of laser intensity with depth (greater than 4 to 1). These data are then processed in typical fashion as well as in by artificially constraining the depth to less than half to mimic the effects of coaxial illumination on the signal to noise ratio of the data (going from 4:1 to 1:1). It was expected that tomographic reconstruction would fail in such a case, but the study found that while the tomographic reconstruction was affected, the eventual velocity fields had relatively small error compared to other sources of error such as misalignment of the cavities. These results suggest that coaxial illumination may be feasible for Tomo-PIV in some cases. Separately, in many hydrodynamic turbulent shear flows, the weaker, stream-wise oriented vortices will cavitate before the stronger, span-wise vortices. This occurs due to stretching of the weaker vortices by the stronger vortices, which leads to a reduction in the core pressures of the weaker vortices. The stretching leads to reduced vortex core diameters and axial flow, but the relative interplay of these was previously unclear. This study experimentally examines a canonical case of this vortex interaction and inception process by looking at a pair of initially parallel line vortices undergoing the Crow instability. The vortices are generated by hydrofoils (modified NACA 66 profiles with chord 167 mm and varying span) whose relative arrangement can be adjusted to vary the flow properties. The vortices were characterized over a range of speeds, from 1 to 10 m/s, with Shake-the-Box particle tracking velocimetry (STB) using the same laser and cameras as above. STB measurements at 3 m/s throughout the downstream region were used to infer pressure drops in the cores of the vortices, with typical pressure drops in excess of 100 kPa found. Reduction in core radius of the stretched vortex and not axial jetting was found to cause the pressure drop. Inception experiments with acoustic measurements over a range of freestream pressures (70 to 150 kPa absolute) at fixed speed of 10 m/s were then done to find inception event rates (approximately 0.01 – 10 events/sec). The inception event rates were predicted reasonably by combination of pressure drops from STB and measurements of nuclei in the water. The acoustic emissions of inception are related to the nuclei and vortex flow properties.