(149c) A Comparative Study of Euler-Euler and Euler-Lagrange Mesoscale Simulations of Moderately Dense Gas-Solid Flows
Accurate modeling of the dynamic behavior of gas-solid flows such as bubbling and circulating fluidized beds is important for the chemical and energy industries. Although Euler-Lagrange (EL) approaches have gained considerable popularity for modeling gas-solid flows in the past decade, the computational cost of fully resolved EL simulations is typically very high. Consequently, Euler-Euler (EE) approaches, such as kinetic-theory-based two-fluid models (TFM), remain the major workhorse in this area. However, the underlying assumption in TFM that the anisotropic part of the particle-phase pressure tensor depends on the deformation rate tensor has been proven invalid by many experiment data and EL simulations, especially when particles are dilute or moderately dense. In our previous work, the EE Anisotropic Gaussian approach (EE-AG) has been shown to produce good agreement for key statistical measures with EL simulations - such as particle settling velocity, volume fraction probability density function, two-point cross-correlations - when particles are dilute (1% by volume). In this work, a novel EE-AG/TFM hybrid solution algorithm, based on different contributions to the particle-phase spatial fluxes, is used to conduct fully resolved mesoscale simulations of cluster-induced turbulence when particles are moderately dense (10% by volume). This new algorithm is implemented in an open-source CFD package, and fully resolved simulation results are compared with EL simulations. The results demonstrate that this new approach of modeling in an EE framework can accurately capture the detailed dynamics of the gas-solid flows and produce results comparable to the EL approach at much lower computational cost.