Leveraging event detection techniques and cross-correlation analysis for phase-detection probe measurements in turbulent air-water flows

Abstract

Dual-tip phase-detection intrusive probes are commonly used to characterize gas-liquid flow velocities in hydraulic structures. Traditional post-processing is based upon the cross-correlation analysis of the raw voltage signals of the two probe tips, providing a time-averaged velocity for the full sampling duration. A recent advancement of the traditional signal analysis, the adaptive window cross-correlation (AWCC) technique, allows the computation of pseudo-instantaneous velocities for small groups of particles. A different analysis approach is the particle event detection technique, which has the potential to directly estimate instantaneous particle velocities. Event detection (ED) techniques have received little attention in highly aerated flows and this work provides a first proof-of-concept. Air-water flow data on a stepped spillway have been re-analysed and it was found that a novel particle event detection technique was able to provide reliable particle velocities and velocity fluctuations across the full air-water flow column, with higher uncertainties close to channel invert linked to the three-dimensionality of the flow. A comparative analysis of the event detection technique and the adaptive cross-correlation method showed good agreement in mean velocities and velocity fluctuations, suggesting that event detection techniques may provide a valuable alternative.