Gas cavities, natural and ventilated, can occur on ship propulsors, control surfaces and hulls. These cavities can be significant sources of small bubbles in the ship wake. We present a two-dimensional time resolved X-ray densitometry system developed for investigation of natural, ventilated and mixed cavities. First we consider the limitations and performance of our X-ray system, and compare the measured void fraction to stationary known void fraction, and to data for a ventilated cavity obtained using dual fiber optical probes. Second, we present preliminary data from time-resolved X-ray used to observe the overall dynamics and time dependent void fraction distribution of a cavitating backward facing step with and without gas injection. © 2013 - IOS Press and the authors. All rights reserved.

The mass transfer through gaps connecting two adjacent channels was investigated as a function of gap geometry and flow conditions. An experiment with a simplified geometry was conducted to aid in the physical understanding and to provide data for validation of numerical computations. The flow loop consisted of two channels with two interchangeable test sections. The inlet Reynolds number in each channel could be independently varied from 4xl04 up to 1x10s. Measurements were performed for seven channel flow rate combinations and eleven gap heights for both test sections. The mass transfer through the gap was calculated from mass flow rate and tracer concentration measurements taken at channel inlets and outlets. Planar and tomographic particle imaging velocimetry, as well as imaging of fluorescent tracer dye, were utilized for select conditions to examine the dynamics of the mixing. Accompanying computations were performed and the results compared favorably with experimental data. For the cases of nearly balanced flow, large coherent structures forming in the gap were observed and exhibited a normalized frequency in agreement with that reported by previous investigators. Over the tested range, the mixing rate as a function of gap height was nominally independent of channel Reynolds number. For significantly unbalanced flow the measured mass transfer approached the one-way mass transfer limit, whereas for larger gaps and closer flow balance the mixing due to coherent structures became significant.

Two-phase mass transfer through a narrow gap connecting two adjacent channels was investigated as a function of gap geometry and flow conditions. The vertical test section consisted of two 127 mm × 127 mm channels connected through a 1,219 mm (L) × 229 mm (W) height-adjustable gap (0-50 mm). The single-phase (water) inlet Reynolds number for each channel was independently varied from 4×104 to 1×105. The gross single phase fluid exchange between the flow channels through the connecting gap, or mixing, has been previously characterized. For the two-phase experiments, air was injected to either or both flow channels inlets via a needle array to produce nominally monodispersed bubbles with a mean diameter of 5 to15 mm, depending on the air flow rate. The air flow rates were metered at the inlet and varied to achieve a crosssectional void fraction of 1% to 15%. Multi-phase mixing through the gap was quantified based on the measured mass flow rates of the water and air and through measurement of a liquid dye tracer concentration at the inlet and outlet of each channel. The void fraction, bubble size, and gas phase velocity were measured using dual-plane wire mesh conductivity sensors at both inlets and outlets. Synchronized multi-view imaging of the fluorescent tracer dye and air bubbles provided visualization of the mixing phenomena. A direct comparison of the single- and multi-phase mixing coefficients showed that the fraction of leakage between the channels could be reduced by more than 80% by the addition of air bubbles to the channel flow. The integral mixing coefficients varied with the relative volumetric flux of liquid and gas. Modification of the single-phase mixing, due to the presence of the air bubbles is discussed.