(452h) Simulation of gas-solid flows in large-scale systems using the direct simulation Monte Carlo method

In this work, the direct simulation Monte Carlo (DSMC) method is developed as a framework for simulating large scale systems with complex physics and hydrodynamic phenomena. While the discrete element method is a Lagrangian approach that can readily incorporate complex particle-scale physics, its use is limited to relatively small systems because of computational costs. Continuum methods such as the two-fluid model can simulate large-scale systems, but there are challenges incorporating particle-scale physics in an Eulerian framework. The DSMC method is a Lagrangian approach that retains the ability to incorporate complex particle-scale physics but is more computationally efficient than discrete element methods. The DSMC framework tracks computational particles that represent many real particles, and the representative particles can experience the same dynamics as real particles. Collisions between computational particles are modeled based on the kinetic theory of granular gases, and the DSMC approach stochastically solves the collision integral of the Enskog equation. Therefore, the DSMC method does not require simplifying assumptions that are often used to derive continuum closures, such as closeness to equilibrium and small Knudsen numbers. To validate the DSMC implementation, we show that the method accurately predicts the onset of the clustering instability in a homogeneous granular gas. The kinetic theory of granular gases breaks down at high solids concentration because of enduring multi-body contacts. To overcome this limitation, the DSMC method is coupled to a deterministic discrete parcel method when the granular flow transitions to the frictional regime. The DSMC approach is then used to model a fluidized bed and industrial scale riser. The simulation space spans the dense to dilute flow regimes and the results are compared to discrete element simulations.

Keywords: Particle technology and fluidization, computational fluid dynamics