Abstract
Process equipment is widely used in industry today and applies complex physical interactions when used in simulations. A need has emerged to obtain a fully working and accurate simulation of process equipment involved with multiphase flows. Through research and application of various numerical theories to multiphase flow simulations this can be achieved. Thus, the aim of this study is to achieve a fully developed and stable air-core model with particle separation whilst implementing the Eulerian-Eulerian multiphase model and the Reynolds Stress Model (RSM). Focus will be placed on ensuring model accuracy when compared to experiment as well as aircore formation and stability. A brief literature study was conducted in order to gain insight into the most recent stage of development in hydrocyclone modelling. A physical experiment was conducted in order to obtain the boundary and input conditions for the simulation such that the results obtained from the simulation can be experimentally validated against the experimental results. Silicon (MQ 15 Crystalline Quartz) particulates and a hydrocyclone barrel diameter of 50mm were used in the experiment. These conditions remained the same when the simulations were conducted. Two air-core models were constructed with different turbulent models and the particulates model was constructed as a continuation from the air-core models. The first air-core model utilised the Eulerian-Eulerian approach coupled with the Renormalization Group (RNG, k-ϵ) model theory. The second air-core model also utilised the Eulerian-Eulerian approach but with the RSM. These models were conducted first in order to reduce computational instabilities that may occur when adding particulates to the system. Comparisons between the two models were made and were also validated against experimental results. Both achieved fully developed and stable air-cores and accurate results of predicted mass flow rate at the overflow in comparison to experiment. Both models did not produce accurate results when predicting the mass flow rate at the underflow. The particulates model was analysed regarding air-core formation, particle separation and particle physics. The results from simulation produced a partially developed aircore with accurate results in predicting the mass flow rate at the overflow in comparison to experiment. The predicted mass flow rate at the underflow was noticeably over-predicted in comparison to experimental results. The particle separation efficiency was found to be fairly accurate in comparison to experimental results. The challenges experienced with the particle simulation results were deemed to be the result of insufficient run time and the lack of an adequate particle interactions model. The results from the study showed that a fully developed and accurate Eulerian- Eulerian based air-core model can be achieved although difficulties arise when incorporating particles into the system due to the complex physical interactions taking place especially at the underflow. Incorporating more complex numerical models such as the coupled Computational Fluid Dynamics (CFD) – Direct Element Method (DEM) model with the Dense Discrete Phase Model (DDPM) for additional granular interactions may improve the mentioned problem areas.
M.Ing. (Mechanical engineering science)