Calibration of a DEM contact model for wet industrial granular materials

TL;DR

The paper develops a DEM framework to model wet industrial polypropylene powders in the pendular regime by incorporating capillary bridges and a dynamic liquid-migration model. It calibrates two physically interpretable parameters, V_min and V_max, against dynamic angle of repose data from rotating drum experiments for two powders with differing porosity and morphology, demonstrating successful reproduction of bulk flow for PP2 across moisture levels and for PP1 up to moderate moisture. The study highlights the utility of dynamic AoR as a calibration target while acknowledging scaling artefacts that emerge at higher liquid contents due to agglomeration, and it provides practical guidance for applying DEM to wet powder processing. It also outlines future directions to reduce scaling artefacts, extend beyond pendular states, and develop agglomeration-aware scaling laws for more accurate and efficient simulations in industry.

Abstract

This study presents and calibrates a Discrete Element Method (DEM) contact model for wet granular materials in the pendular regime. The model extends a previously calibrated dry contact formulation by incorporating liquid bridges that generate capillary adhesion between particles, while liquid migration is represented through evolving bridge volumes. Two reactor-grade polypropylene powders with different particle size distributions, bulk densities, and surface morphologies are investigated, resulting in distinct wetting behavior. A schematic framework is introduced to relate increasing liquid content to the transition from dry to wet contacts using two key parameters: the minimum liquid film volume and the maximum liquid bridge volume. These parameters are calibrated using dynamic angle of repose measurements from rotating drum experiments. The calibrated model reproduces the experimental flow behavior of both powders: full agreement is obtained for the coarser, more porous powder across all liquid contents, while for the finer, denser powder, agreement is achieved at low to moderate liquid contents. At higher liquid contents, discrepancies arise due to agglomeration effects amplified by particle scaling. These results demonstrate the effectiveness of the dynamic angle of repose as a calibration target and highlight the limitations of particle scaling for strongly cohesive wet granular systems. The proposed framework provides a practical basis for DEM-based modeling of wet powder flow in industrial processes.

Figures (15)

• Figure 1: Schematic illustration of the Innovene polypropylene (PP) process, adapted from dittrich2007residence.
• Figure 2: Illustration of wetting states as a function of liquid content, adapted from mitarai2006wet.
• Figure 3: Illustration of the formation (left) and rupture (right) of a liquid bridge. Blue particles are wetted ($V_f > 0$), gray particles are dry ($V_f = 0$), and blue lines represent liquid bridges.
• Figure 4: Experimental dynamic angle of repose ($\theta$) for PP1 and PP2 as a function of liquid content.
• Figure 5: Experimental snapshots of PP1 and PP2 at different liquid contents, showing the flow behavior.