(110c) A Multi-Regime Continuum Approach for Modeling Flow of Granular Pine Chip

The flowability issues in biorefineries are due to poor understanding of the flow behavior of biomass granulate material, which leads to inappropriate equipment design and operation. With overall objectives of providing design tools and operational envelopes of each handling equipment for biomass-derived feedstock (ground pine chip), this talk will present the computational modeling efforts to address these issues by combining the discrete element method (DEM) at the particle scale and the Eulerian mesh-based finite element method (FEM) at the continuum scale. At the particle scale, we evaluate and identify the most suitable DEM model to efficiently and accurately realize the shear behavior of ground pine chip with complex particle size & shape distribution and particle compressibility. X-ray CT scan data and sieve analysis are used to generated DEM particle morphology, axial shear experiments are then used to calibrate and validate the contact model parameters. The high-fidelity DEM model with calibrated parameters is used to generate ground pine bulk flow behavior (stress-strain rate relation), and then up-scaled to evaluate and formulate the constitutive model. At the continuum scale, the critical state theory based NorSand model, incapable of modeling the nonlinear elasticity, the pressure- and void ratio-dependent shear flow, softening and hardening accompanied by volume compaction and dilation, is evaluated and identified as the most suitable constitutive model to capture milled biomass flow at the quasi-static condition (solid-like). We enrich the NorSand model with dynamic modeling capability by incorporating a rheology model, in which the shear strength of granular material is dependent on shear rate and pressure (fluid-like). We calibrate the constitutive model parameters controlling quasi-static behavior from cyclic uniaxial compression and Schulze ring shear test, and we obtain the constitutive model parameters controlling fluid-like behavior from inclined plate flow experiments and DEM generated stress-strain rate relationship. After implementing the extended model with the capability to model multi-regime flow from quasi-static to dynamic into the FEM package, it is utilized to model the hopper flow with the same geometry and initial conditions as the experiments at the pilot-scale. The predicted mass flow rate with the variation of hopper inclination angle and of the opening size is compared to experimental data to validate the model. We further perform sensitivity analysis by using the new multi-regime model and its calibrated parameters to search for the optimal design of hopper at the industry-scale.