(110f) Computational Fluid-Dynamics Modeling of Biomass-Feedstock Flow Using a Non-Local Granular-Fluidity Constitutive Model

Upsets in milling, material handling, and reactor feeding systems significantly contribute to the downtime in biorefineries. We are working towards reducing these upsets by developing constitutive relations derived from the physical attributes of biomass feedstocks to couple with Computational Fluid Dynamics (CFD). In particular, here we present work on using the Non-Local Granular Fluidity (NLGF) constitutive model to represent dynamic flow of bulk biomass feedstocks. Whereas other constitutive models, such as Drucker-Prager-Cap and inertial rheology, are generally capable of resolving bulk flow in many flow regions, they are not capable of representing the well-known nonlocal phenomena in granular flow. Nonlocal phenomena emerge in regions where flow features occur on length scales that are near the scale of the system geometry, e.g., the jamming of hopper outlets and the stop-height of piles on inclines. The NLGF model handles the nonlocality of flow by introducing an intrinsic diffusive length scale into the granular rheology model. The developed CFD model, based on NLGF, was used to simulate biomass feedstock flow in feed-handling equipment and into reactors. Toward this goal, we extend the existing software packages and algorithms to accommodate the computational modeling work. Further, we validate the NLGF model-based CFD simulations of biomass feedstock flow by comparing simulation results against lab-scale flow and deformation experiments. Specifically, the flow of particulate pine residues in three systems will be presented: inclined plane setup, ring-shear tester, and a wedge hopper. This experimentally validated model will be used for providing actionable guidance for operating and optimizing lab to pilot scale biomass-conversion feeding systems.