Magnetic Resonance Velocimetry for Multiphase Flow Mapping: Probabilistic Identification of Time-Averaged Bubble-Rich Zones

Magnetic Resonance Velocimetry (MRV) enables non-invasive, three-dimensional, three-component velocity field measurements in complex flow environments. While MRV has been widely applied to single-phase flows, its use in multiphase systems remains limited due to MRI’s reliance on hydrogen nuclei. Recent efforts have attempted to visualize multiphase flows by interpreting MR signal loss caused by air bubbles. However, such signal loss may also originate from turbulence or rapid acceleration, complicating the interpretation—particularly in high-Reynolds-number flows. In this study, we explore the extension of MRV to visualize three-dimensional, fast-moving multiphase flows, with an emphasis on identifying time-averaged bubble-rich regions. Recognizing that the presence of air bubbles introduces localized random noise in the measured velocity field, we propose a probabilistic interpretation of this noise to distinguish bubble-induced signal loss from other physical effects. This approach was applied to the sump region of a dishwasher—a practical example of turbulent multiphase flow—where MRV captured complex three-dimensional velocity structures. From these data, we probabilistically identified time-averaged bubble-rich regions and estimated the surrounding flow patterns near the internal filter. Further results and discussion will be presented at the conference.