Experimental and numerical investigation of die designs in biomass pelleting and the effect on layer formation in pellets

Simon Klinge Nielsen, Matthias Mandø

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

Design parameters of a pellet mill die for pellet production are essential for running optimal pellet production in terms of energy consumption and quality of the pellets. In this study, the effects of the countersink angle and depth are investigated through experimental tests and Computational Fluid Dynamics (CFD) simulations. Inlet die designs with angles 0°, 60°, and 100° and three different depths, were tested via single pelleting tests with spruce. A new design parameter, AR, is suggested for comparison of the dies performances. The parameter is derived from the die's surface area, and is the ratio of the die's inlet area vs. total surface area and values tested here ranged 0.35 to 1.

Specific energy consumption and mechanical pellet durability were measured experimentally, and the feedstock layer profiles in pellets were qualitatively analyzed via image processing. The layer profiles in the pellets were simulated with a CFD model using a simple Bingham viscosity model. The model was validated by comparing the simulated and experimentally obtained layer profiles. The results show that the lowest energy consumption was obtained with a 60° countersink and an AR value of 0.6–0.8. Furthermore, an AR value in this range were found to be optimal with respect to pellet durability. Analysis of the layer profiles shows interesting differences in the layer profiles of the pellets, and evaluation of the kurtosis of the layer profiles suggests that this can be used to predict the pellet durability and, thereby, the quality of the die design.
Original languageEnglish
JournalBiosystems Engineering
Volume198
Pages (from-to)185-197
Number of pages13
ISSN1537-5110
DOIs
Publication statusPublished - Oct 2020

Keywords

  • Biofuel
  • Image processing
  • Pelleting
  • Rheology
  • Simulation

Cite this