Non-linear dynamics of fluidized beds: Insights from coupled CFD-DEM simulations

Abstract

Accurately capturing the highly nonlinear gas-solid interactions within fluidized beds remains a formidable challenge, critically hindering their robust design and optimization for industrial applications. Conventional numerical characterization methods often focus only on time-averaged data, failing to resolve the inherent dynamic complexities governing these systems. This study proposes an integrated approach, combining Computational Fluid Dynamics - Discrete Element Method (CFD-DEM) simulations with advanced nonlinear dynamics analysis, to comprehensively elucidate the true dynamic characteristics of both monodisperse and polydisperse gas-solid fluidized beds.We systematically investigated dynamic pressure and solids distribution fluctuations extracted from high-fidelity CFD-DEM simulations across varying superficial gas velocities, compositional variations, probe positions, and aspect ratios. Employing tools from nonlinear dynamics theory, this analysis rigorously unveiled the intrinsic nonlinearity of fluidized beds, confirming their profound sensitivity to initial conditions and complex, unpredictable behavior. Specifically, Kolmogorov Entropy, quantifying information loss and system unpredictability, and Correlation Dimension, measuring system complexity and degrees of freedom, were precisely quantified. These robust metrics provided unprecedented, nuanced insights into fluidization regime transitions, the critical influence of polydispersity on dynamic behavior, the intricate dynamics of bubble formation and interaction, and the precise identification of highly chaotic regions within the bed.This study significantly advances the understanding of fluidized bed dynamic characteristics by introducing a robust methodology that transcends conventional reliance on averaged properties. The integration of CFD–DEM with nonlinear dynamics analysis demonstrates high efficacy, yielding robust, data-driven insights crucial for optimizing reactor design and enhancing operational control. Furthermore, by successfully capturing dynamic profiles, this combined approach establishes a strong foundation for rigorous investigations into the scale-up of fluidized bed reactors, delineating a clear trajectory for future research.