Automated Optimization of Mixing Elements for Single-Screw Extrusion Using CFD Simulations

Today’s materials must meet high mechanical requirements while remaining cost-effective in production. This requires a homogeneous temperature and material distribution and splitting, as achieved by mixing elements in single-screw extrusion, which depends on the material properties, the geometry of the mixing elements, and the process conditions. Existing computational fluid dynamics (CFD) methods can help, but often optimize dispersive (split) or distributive (spread) mixing separately and neglect their mutual influence and competition, which prevents a single optimal solution. To address this, this work develops an automated optimization tool using a genetic algorithm for the holistic optimization of both mixing processes, considering pressure drop, temperature gradient, and quantitative metrics for dispersive and distributive mixing. Compromises between geometry and metrics that improve dispersive mixing while maintaining moderate temperature gradients and pressure drops were determined for four different polymers. The results of the dispersive mixing element show dependencies between mixing metrics and geometry. In contrast, the distributive mixing element shows no clear correlations between mixing metrics and geometry.