CFD modelling with MFIX® of liquid-solid fluidized beds of polydisperse struvite crystals

Abstract

This final master project aims to model with computational fluid dynamics a fluidized bed of polydisperse struvite crystals to analyze, study and compare the fluid dynamic properties of the bed against the results that (Md. Saifur, et al. 2017) show in their article (CFD modeling of liquid-solid fluidized beds of polydisperse struvite crystals) published the 21 October of 2017. A steady state is reached through a transient study of polydisperse solids initially packaged. These solids are fluidized at different up-flow velocities (different case studies). The validation is going to be done once the steady state is reached at the real time of 120 s. Also, the characteristic fluidization heights reached in the steady state are evaluated for each set of solids of a determined diameter. The validation data between the simulation and the pilot plant (literature) is the time averaged liquid volume fraction (also known as bed voidage and being the complementary value to the solids fraction). Each MFIX® used module is described. The explained modules are the geometry design, FAVOR (fractional area/volume orthogonal) mesh generation, region generation and the boundary and initial conditions. ParaView® open code program is used to analyze, interpret and study the exported results. The data is shown in vector and contour picture scale and in numerical-graphical level for each exported time step data. Also, videos that reproduce the virtual reality of the study are exported from ParaView®. The chosen geometry has been 2D, in order to reduce the computational time and to adapt better the FAVOR mesh to the studied region. The bed dimension is a cylinder with 1392 mm high and 100 mm diameter. The pilot plant was done in a 20 ºC temperature-controlled room, thereby this value is constant in all the studies. The system outlet pressure is 1 atm (boundary condition for all case studies). The bed inlet velocity is a uniform inlet up-flow velocity, the value of this boundary condition is changed for all different case studies.The values predicted by both ANSYS® and MFiX® CFD simulated models are in excellent agreement with the pilot plant experimental data obtained by (Md. Saifur, et al. 2017).