Multiscale modelling of cavitating flow with considering non-condensable gas nuclei

Cavitation, a critical phenomenon in fluid dynamics, significantly impacts hydraulic machinery and propellers, causing performance loss, vibration/noise, and material damage. While traditional research focuses on pure-liquid cavitation, real-world fluids often contain non-condensable gas nuclei (e.g., microbubbles or dissolved gas), which critically influence cavitation inception and dynamics. This study introduces a multiscale Eulerian-Lagrangian framework to model cavitating flows with non-condensable gas nuclei. The Eulerian Volume of Fluid (VOF) method resolves large cavity interfaces, while the Lagrangian Discrete Bubble Model (DBM) tracks microbubble dynamics. This hybrid approach captures interactions among vapor cavities, vapor bubbles and gas nuclei, improving simulations of sheet and cloud cavitation. Numerical results for cloud cavitation show strong agreement with experimental holographic imaging data, particularly in cavity shedding and wake flow regions, validating the model accuracy in predicting bubble number densities (vapor and non-condensable gas bubbles). The findings enhance understanding of gas nuclei effects on cavitation evolution in practical applications.