Mixing performance for wet and sticky bulk materials in a vertical blender using DEM simulation

Abstract

Wet and sticky bulk materials exhibit poor flowability during the mixing process, which prevents adequate contact between dry and wet particles. This results in uneven moisture distribution and deterioration in the mixing system. To address these issues, the mixing process of viscous concrete was focused on a vertical blender. A comprehensive investigation into mixing mechanisms and particles flow patterns were conducted using the discrete element method (DEM). The accuracy of the contact parameters in DEM was calibrated through repose angle and validated by torques tested in a custom-built mixing platform. And then, the effects of moisture content, filling level, rotational speed, and inclined angle were systematically investigated with respect to key mixing metrics: the relative standard deviation (RSD), coordination number (CN), segregation index (SI) of wet particles, as well as liquid mass. The results indicated that when the moisture content is 8 %, filling level is 50 %, and rotational speed is more than 60 rpm, the CN and mixing efficiency are acceptable, and the RSD and SI are low, thereby improving the mixing quality. The convective motion was revealed as the dominant flow regime through statistical quantification of diffusion coefficients and Peclet numbers. Finally, Box-Behnken Design was employed to develop quadratic polynomial models for RSD, CN, and SI, which demonstrated strong accuracy in predicting mixing performance and enabled systematic optimization of critical process parameters.