Modelling Single Particle Settlement by CFD-DEM Coupling Method

Abstract

Dust liberation (a gas-solid two phase problem) is a highly complicated phenomenon to be studied due to great number of variables to be considered, so any related theoretical or numerical simulation is forced to implement many simplifications. Furthermore, there are numerous methods to study such a problem one of them is the newly introduced CFD-DEM coupling method. This method is expected to be used extensively in near future; therefore, it is essential to have an understanding about its applicability, strong points and shortcomings to be able to model complex problems such as dust liberation. Accordingly, this research is aiming to model single particle settlement (SPS) as the simplest version of dust liberation problem using CFD-DEM coupling method to comprehend both the SPS problem and the method. To reach this goal three main steps has been taken. In the first step the focus is on understanding the single particle settlement problem. In settlement of a single particle two sorts of forces play roles, driving forces and damping forces. The only driving force is the force of gravity; damping forces, which are opposing the gravity force, are buoyancy, drag, virtual mass, and Basset and lift forces. Results indicate that buoyancy as well as drag force are most influential damping forces. Basset and virtual mass forces are important in the transient phase of the problem only and if the ratio of fluid density to particle density is smaller than a certain value (10-3), the influence of these forces in transient phase as well as steady phase is negligible. In the second step, reasons why CFD-DEM coupling method is preferred for modeling SPS is elaborated. Based on the final application of this thesis, which is modelling of dust liberation phenomenon for research purposes, three criteria are formulated to identify the suitable method. Firstly, the method should be able to study the microscopic mechanisms; secondly, it should be capable of handling particle-particle and particle-wall collisions. Finally, the needed computational effort should be at least in acceptable range. Considering these criteria, the CFD-DEM coupling method has been chosen because of its ability to study microscopic mechanisms, its capability in capturing the particle interactions, and its acceptable computational effort. In the final step, the performance of CFD-DEM coupling method is evaluated. To do so, the numerical results have been compared to experimental and analytical results. In the calibration step, the influential parameters such as governing equation model, void fraction model and drag models have been studied and the most appropriate ones have been chosen. In the verification step, results of the numerical model are compared to other experimental cases with different fluid properties. Consequently, in the phase of validation, results of the numerical model are compared to analytical results of another problem with different particle size, geometry and fluid. Based on the observations, it can be concluded that CFD-DEM coupling method can perform accurately and efficiently in the steady-state phase of problems by considering the following two points. Firstly, the void fraction model choice should be mainly dependent on particle diameter to container size ratio. Secondly, the improved big particle void fraction model should be used for simulations with fine mesh and the smoothing length is suggested to be in particle size range.