Total: 1
The Archimedes Tube Crystallizer (ATC) employs air-segmented flow in coiled tubes to achieve narrow residence time distributions for continuous crystallization. Taylor and Dean vortices drive particle suspension in this system. However, one-way coupled models fail to capture the fluid-particle feedback that becomes critical at higher loadings. We present a particle-resolved Direct Numerical Simulation (DNS) framework based on a Finite Element-Fictitious Boundary Method with hard-contact modeling of particle interactions. Simulations of L-alanine suspensions across varying particle sizes, solid contents, and rotational speeds are validated against experimental side-view imaging. Three quantitative metrics-axial distribution, radial index, and vertical asymmetry-are introduced to classify suspension regimes. The DNS results reproduce the experimentally observed flow map zones (green, yellow, red/yellow, red) and resolve subtle transitions such as rear loading and loss of vertical symmetry. This feasibility study demonstrates that DNS can reliably predict dense suspension behavior and provides a mechanistic foundation for crystallizer design.