CFD modelling and multi-objective optimization of MHO for hydrodynamic cavitation generator using a radial basis function neural network, and NSGA-II

In this study, the design of a multi-hole orifice (MHO) commonly used in cavitation reactors, is optimized to maximize the vapor production and the turbulence intensity downstream of the MHO. The independent opti-mization parameters are non-homogenous distribution for peripheral holes (vertical radius, & delta;y, and horizontal radius, & delta;x), the diameter of the central hole, n, and the orifice thickness, n. A design of experiment, a set of CFD simulations, and radial basis function neural network (RBFNN) are used to study the effect of the mentioned independent parameters on the two-objective functions. The model analysis shows that the orifice thickness is a most influential factor in vapor production, while the diameter of the central hole remarkably influences the intensity of turbulence following the MHO. The, NSGA-II algorithm is used for obtaining Pareto optimal design, which propose that an MHO with a relative orifice thickness around n = 0.15 produces both maximum vapor output and remarkable turbulence intensity. The turbulence intensity is maximized by a nonhomogeneous dis-tribution of holes, i.e., & delta;x = 0.4 and & delta;y = 0.6. Therefore, & delta;x and & delta;y are responsible for collapsing bubbles due to jets mixing behind MHO. As compared to MHO with n= 0.05, MHO with n= 0.2 showed a pressure recovery length that was twice as long.