Mohammad-Reza Erfanian

Ventilated supercavitation, a complex two-phase flow, has mostly been explored through experiments and simulations, with machine learning yet to emerge as a complementary research approach. This study combines experiments, numerical simulations, and machine learning models to explore the characteristics of ventilated supercavitation and the different cavitating regimes behind a disk-shaped cavitator. The experimental data were used to validate the simulation models. Subsequently, an optimized Random Forest model, enhanced using the Bayesian Optimization Algorithm (BOA-RF), was trained on the simulation results to predict cavitation length, cavitation number, and classify the cavitation regime type. The BOA-RF model exhibited high accuracy when compared with both experimental and numerical results. The results indicate that the supercavity length increases with the ventilation coefficient (CQ). However, at the formation ventilation coefficient (CQf), the cavity length undergoes a significant and rapid increase. Moreover, as the ventilation coefficient (CQ) increases, the cavitation regime evolves from foamy cavitation to a continuous transparent and asymmetric cavity, and when CQ ≥ CQf, it transitions to clean supercavity. Additionally, the results indicate that as the Froude number increases, the CQf initially increases and then decreases.

Ventilated supercavitation is a hydrodynamic phenomenon where injected air forms a large bubble around a high-speed submerged body, reducing drag to enhance its movement. The study of cavitating flows is crucial near the free surface due to significant interactions between the supercavity and surface waves. In this study, the behavior of the ventilated supercavity and free surface waves at high Froude numbers and varying submersion depths has been investigated using a numerical simulation approach. The accuracy of the predicted results is verified by comparing them with the available experimental data. The results indicate that the supercavity size is simultaneously affected by submersion depth and Froude number. Additionally, as the Froude number increases, the submersion depth at which the free surface effect on the supercavity shape becomes negligible also increases. Moreover, near the free surface, unlike in fully submerged conditions, the supercavity length increases with depth. The results also show that three types of free surface wave patterns are formed, depending on the intensity of the first and second wave systems.

One of the complex challenges in ventilated cavitating flow studies is analyzing hysteresis behavior and the formation air entrainment coefficient (Cqf) under different conditions. This study explores the formation and collapse processes of the supercavity using experimental observations, numerical simulations, and machine learning (ML) models to reveal the hysteresis behavior of air entrainment at different flow conditions. Initially, the research focused on studying air entrainment hysteresis for a disk-shaped cavitator under different Froude numbers (Fr) through experimental and numerical methods. The study identified two key air entrainment coefficients in the hysteresis curves, which are important for gas generator design. In the ML section, hyperparameter optimization for the random forest (RF) model is performed using genetic algorithm (GA) and particle swarm optimization (PSO). The results demonstrate that the GA-RF model is more accurate than the PSO-RF model in predicting experimental data. The GA-RF findings show that, for a fixed cavitator diameter, Cqf increases with Fr, reaching a maximum value (MaxCqf) before decreasing as Fr continues to rise. Additionally, with an increase in the cavitator diameter, MaxCqf increases and Fr belonging to this ventilation coefficient (FrMaxCqf) decreases.

The aim of this paper is to investigate the ventilated cavitating flow behavior, through experimental and numerical methods. The experiments are performed for a test model placed in a water tunnel and a high-speed camera is used to image the formation and collapse of the supercavity. For the numerical simulations, Ansys CFX commercial software is employed to solve the Reynolds-averaged Navier–Stokes (RANS) equations and additional transport equation for the liquid volume fraction in the unsteady condition. Both the experimental observation and numerical prediction show a hysteresis behavior for ventilated supercavity in which the value of the air entrainment coefficient required to maintain the supercavity in the formation process is less than the amount required to form it. The good agreement observed between the numerical prediction and experimental data, revealed the accuracy and capability of the numerical scheme. Also, the numerical simulation shows that the variations in the supercavity length, in addition to the air entrainment, are a function of the air leakage regime from the closure region of supercavity.