PhD Defence Roxana Saghafian Larijani | Accelerated Multi-scale modelling of wet granular systems

Accelerated Multi-scale modelling of wet granular systems

Roxana Saghafian Larijani is a PhD student in the Department of Soil MicroMechanics. Promotors are prof.dr. V. Magnanimo and prof.dr. S. Luding from the Faculty of Engineering Technology.

Wet granular systems are ubiquitous in both natural phenomena and industrial applications. In such systems, the presence of liquid between particles leads to the formation of bridges, and the resulting cohesive forces significantly influence the rheological behaviour of the bulk material. Understanding and predicting the behaviour of wet granular systems is therefore crucial for geohazard, such as landslides and avalanches, mitigation and for the optimisation of industrial processes. However, a deep understanding of these systems requires particle-scale information, which is often difficult to obtain experimentally. Moreover, relying solely on trial-and-error experiments is time-, material-, and energy-intensive. These challenges highlight the potential of physics-based computer simulations using the Discrete Element Method (DEM). DEM is a powerful tool to study and optimise granular systems and provides an alternative to resource-intensive experiments. However, the use of DEM is limited by the high computational cost, due to the large number of particles needed to reproduce industrial-scale problems. Upscaling through coarse-graining (CG), i.e., representing a group of original-sized particles by larger particles while applying essential scaling rules, offers a promising solution to improve the computational efficiency of DEM simulations.

This doctoral thesis, comprehensively investigates the applicability of CG-DEM for wet granular systems through appropriate scaling rules, and leverages this method to enhance process understanding using micro-scale information. In particular, the thesis focuses on wet granulation, a critical process in the pharmaceutical, chemical, and food industries.

After an overview of scaling rules for particle-scale mechanisms, the work validates these rules for wet granular systems, demonstrating that CG-DEM with appropriate scaling effectively balances accuracy and computational efficiency. This foundation enables a novel multiscale model combining CG-DEM with Graph Network Analysis (GNA) to study wet granulation. Finally, the thesis applies DEM to analyse thermal and mechanical behaviour of particles in a rotating drum dryer, linking simulations to experimental observations and direct industrial applications.