Artificial cavitation, a complex two-phase phenomenon, involves cavity formation by injecting air into a low-pressure region behind the cavitator. Influenced by factors like cavitator shape and angle, ventilation coefficient, etc., its understanding is crucial. While past research primarily focused on large ventilation coefficients, this study explores the impact of cavitator angles on ventilated cavity characteristics at low ventilation coefficients. For this purpose, three experimental, numerical and machine learning methods have been used. Initially, three angles (0◦, 12.5◦, 25◦) were experimentally investigated across ventilation coefficients of 0.017–0.085. Additionally, through numerical simulations, the effects of increasing the cavitator angle up to 40◦ at small ventilation coefficients were explored. By optimizing random forest (RF) and Extreme Gradient Boosting (XGB) using Bayesian Optimization Algoritm (BOA), the cavitation characteristics under varying cavitator angles were predicted. The results show the appropriate accuracy of the BOA-XGB model compared to the RF model in predicting the characteristics of cavitation. Results indicate at very small ventilation coefficients, the shape of the cavity is not affected by the cavitator angle. However, with the increase of the ventilation coefficient, in addition to its value, the cavitator angle also affects the shape of the cavity. Increasing the cavitator angle under constant ventilation coefficient induces asymmetry in the cavity’s end, elongates the cavity, and reduces the cavitation number.

Keywords