Optimizing the design and performance of solid–liquid separators
Mohsen Izadi, Abbas Misaghe Makvand, Ehsanolah Assareh, Farzad Parvaz
Abstract
This research sought to examine a solid-liquid cylindrical cyclone separator, using the CFD simulation technique. By simulating the multi-phase flow using the Eulerian–Eulerian approach, the SIMPLE Pressure-velocity coupling solution algorithm was considered to simultaneously perform the pressure and fluid velocity analysis. In this way, we discussed the impact of three different turbulence models (RNG (k-ε), Realizable (k-ε), and RSM) on the results of the simulation, aiming to select the best model with acceptable accuracy among these three models (RSM) to evaluate the two-phase (solid-liquid) flow hydrodynamics. In other words, the present study aimed to optimize the parameters affecting the hydrocyclone separation. For this purpose, in order to conduct a statistical study of the CFD model on hydrocyclone performance parameters, a fluid model was developed based on response surface methodology to simulate target parameters based on design parameters with a mathematical polynomial. Finally, the neural network method was used to analyze and solve the optimization problem.