Multi-Field Comparison of Homogeneous Cavitation Models in a Small Converging–Diverging Nozzle: A Controlled CFD Case Study

Homogeneous cavitation models are widely used in engineering CFD analyses of internal flows, but their predictions may differ depending on model formulation, operating regime, and geometry. This issue is especially relevant in small converging–diverging nozzles, where strong local acceleration and rapid pressure variation may amplify model-dependent responses. This study presents a controlled CFD comparison of two commonly used homogeneous cavitation models, namely the Schnerr–Sauer and Zwart–Gerber–Belamri models, applied to a small converging–diverging nozzle with a throat ratio of 0.06. Three-dimensional steady-state simulations were performed for five inlet velocities from 0.1 to 0.5 m/s under identical numerical assumptions. The comparison included vapor volume fraction, static pressure, and velocity magnitude fields, supported by image-based field-similarity metrics: correlation coefficient, structural similarity index, and normalized root mean square error. The purpose of the study was not to provide an absolute validation or universal ranking of the models, but to assess whether both formulations lead to mutually consistent interpretations of cavitating flow within the adopted numerical framework. The results showed high agreement between the two models for inlet velocities from 0.1 to 0.4 m/s. At 0.5 m/s, however, simultaneous divergence was observed in vapor, pressure, and velocity fields, with the strongest discrepancy occurring for velocity magnitude. The additional similarity metric analysis confirmed that this divergence was not limited to visual differences in vapor contours. These findings indicate that cavitation model comparison in small nozzles should not rely solely on vapor volume fraction, but should include complementary pressure- and velocity-based indicators. Overall, the study supports a multi-field and multi-metric approach for interpreting model-dependent behavior in cavitating internal flow simulations.