DEM modelling for flow of cohesive lignocellulosic biomass powders: Model calibration using bulk tests

• Published in: Advanced powder technology : the international journal of the Society of Powder Technology, Japan Vol. 30; no. 4; pp. 732 - 750
• Main Authors: Pachón-Morales, John, Do, Huy, Colin, Julien, Puel, François, Perré, Patrick, Schott, Dingena
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.04.2019; Elsevier
• Subjects: Chemical and Process Engineering; Cohesion; Discrete element method; Engineering Sciences; Multi-objective optimization; Parameter calibration; Woody biomass powder; Woody biomass powder; Cohesion; Multi-objective optimization; Discrete element method; Parameter calibration
• ISSN: 0921-8831; 1568-5527
• DOI: 10.1016/j.apt.2019.01.003

[Display omitted] •A realistic and calibrated DEM model for cohesive biomass powder is obtained.•A multisphere representation reproduces the elongated shape of particles.•A coarse-graining approach is used to reduce simulations runtime.•Three contact parameters are calibrated using a genetic algorithm of optimization.•Optimal solutions reproduce accurately the experimental physical responses. Biomass feeding problems greatly hinder the industrialization of entrained-flow gasification systems for production of 2nd generation biofuels. Appropriate DEM modelling could allow engineers to design solutions that overcome these flow problems. This work shows the application of a DEM calibration framework to produce a realistic, calibrated and efficient material model for lignocellulosic biomass. A coarse (500–710 µm) and a fine (200–315 µm) sieving cut of milled poplar were used in this study. The elongated shape and the cohesive behavior were respectively simulated using a coarse-grained multisphere approach and a cohesive SJKR contact model. Measurements of three physical responses (angle-of-repose, bulk density, a retainment ratio) allowed calibration of the sliding (µs) and rolling friction (µr) coefficients and the cohesion energy density (CED). Using a statistical analysis, the most influential calibration parameters for each bulk response were identified. A Non-Dominated Sorting Genetic Algorithm was used to solve the calibration multi-objective optimization problem. Several sets of optimal solutions reproduced accurately the three physical responses and the experimental shear responses were closely reproduced by simulations for the population of coarse particles. The DEM calibration framework studied here aims to produce material models useful for assessing flow behavior and equipment interaction for biomass particles.