Evaluation of Turbulence Model on Transient Cloud Cavitation Behavior and Cavitation Erosion Risk in Nozzle Flows

This study systematically evaluates the computational accuracy and efficiency of various turbulence models for predicting cloud cavitation shedding and erosion risk in fuel nozzles. The models investigated include the Reynolds-averaged Navier-Stokes (RANS) model, large eddy simulation (LES), delayed detached eddy simulation (DDES), and very LES (VLES). Results indicate that the RANS model fails to accurately predict cloud cavitation shedding due to its inherent time-averaging limitations. Similarly, DDES exhibits limited accuracy in capturing shedding dynamics, reentrant jet and high-frequency pressure fluctuation. In contrast, LES accurately resolves the whole cloud cavitation evolution cycle, including small-scale vortex structures and transient pressure fluctuation, showing strong agreement with experimental data. Notably, VLES, based on a scale-adaptive hybrid RANS-LES approach, reduces computational costs by ~30% while maintaining high fidelity in resolving cavitation shedding. Fast Fourier transform analysis of pressure fluctuations further validates the efficacy of VLES , which was then employed to investigate cloud cavitation evolution and erosion risk under different inlet pressures. Increasing inlet pressure intensifies both cavitation intensity and pressure fluctuations while decreasing shedding frequency, concomitantly causing reentrant jet disturbances and the erosion risk zone to expand downstream. High cavitation erosion risk regions remain prevalent at super cavitation regime due to enhanced adverse pressure gradients and sustained localized bubble collapse. This study provides theoretical insights into nozzle cloud cavitation flow and erosion mechanisms, offering valuable guidance for nozzle design optimization and erosion risk prediction.