Answers are grounded in this paper and cite their sources.
Cavitation—and, in particular, cloud cavitation—plays a critical role in the operation of fluid systems due to its strong influence on hydrodynamic performance and material erosion. In this study, we present a novel image-processing framework based on a modified U-Net architecture for the detection of cavitation structures from high-speed photography. The model performs pixel-level segmentation of cavitating regions, followed by a post-processing step employing a heuristic algorithm to distinguish between the cavity sheet and the shed cloud. To the best of our knowledge, this is the first framework specifically designed to systematically identify and differentiate sheet and cloud cavitation, a distinction that is essential for many studies in cavitation research. The method is applied to cloud cavitation developing about the benchmark NACA 0015 hydrofoil at a fixed Reynolds number and incidence, while varying the cavitation number. The proposed approach enables the extraction of characteristic length scales, including the unsteady sheet length, cloud size, and their spatial distribution. These measurements provide the basis for a statistical analysis of cavitation dynamics. The framework offers a fast, robust, and versatile tool for detecting cavitation phenomena, providing detailed information on the location, extent, and evolution of cavitation structures. As such, it constitutes a practical tool for the systematic investigation of cavitation phenomena in fundamental research.
No takes yet. Share an insight, caveat, or question.
Hatzissawidis et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: