Simulation-based characterization of alginate aerogel packed bed compaction via DEM-BPM

Abstract

The global demand for aerogels is constantly growing, thus, optimizing and scaling up the production processes have become increasingly important in the last decade. The utilization of millimeter-sized aerogel particles for such purposes is typically preferred due to inherent advantages in handling and production compared to other geometries. The production of these particles is most commonly accomplished using a particle packed bed (autoclave). This process presents, however, several challenges, including the impact of mechanical loads on the quality of the product. Therefore, this work focuses on deepening the understanding of mechanical properties and deformation mechanisms of aerogel particles in packed beds under uniaxial compaction. The investigated alginate aerogel particles are characterized by a spherical shape (circularity of 0.96), a specific surface area of ∼352 m²/g, an average diameter of ∼3.3 mm, and a bulk density of ∼0.05 g/cm³. In addition, this study extends a DEM-BPM model to capture the mechanical deformation of biopolymer aerogels, both as individual particles and within packed beds. The simulations were calibrated and validated using experimental data from uniaxial compaction tests. An optimization methodology was implemented to reduce reliance on traditional trial-and-error methods and improve the model's accuracy. The results demonstrate that the proposed DEM-BPM model effectively replicates the mechanical behavior of alginate aerogels, showing strong agreement between experimental data and minimal deviations for both single particles and packed beds (R²$\ge$ 0.93). This model serves as a promising tool for gaining deeper insights into the mechanical properties of aerogels and improving production efficiency. Additionally, the DEM-BPM model can be expanded to incorporate intermediate products, such as hydrogels and alcogels, enabling process optimization at every stage of aerogel manufacturing.