| The main target of our research is to investigate powder mixing, particularly continuous mixing. Continuous mixing is considered as an efficient alternative to batch mixing processes that in principle allows for easier on-line control and optimization of mixing performance. In order to illustrate the benefits of this process we have demonstrated the effectiveness of continuous mixing for powders. A number of operating and design parameters including processing angle, rotation rate, fill level, convective design, APAP concentration, and residence time have been investigated to consider their effects on mixing performance and on the content uniformity. Statistical analysis has been applied to examine the significance of the effects of processing parameters and material properties on the mixing rate. In addition to mixing experiments, the particle trajectory within a continuous mixer has been studied for different cohesion levels, flowrates, and rotation rates using Positron Emission Particle Tracking (PEPT). The approach was beneficial in providing particle trajectories and, as a result, allowing us to obtain axial dispersion coefficients quantitatively. The experimental methods have been used to verify computational approaches as well as study some important areas that are difficult to examine experimentally such as online homogeneity measurements.;Notably, powder-mixing models are restricted due to computational limitations and obstacles associated with correlating simulation-time to real-time. We have developed efficient modeling approaches that will enable the simulation, optimization, and control of mixing processes. One method is compartment modeling, a method that discretizes the blender into finite regions. We have adapted the approach to mixing processes (v-blender, a horizontal drum, and continuous blenders). Another approach we propose is the use of a hybrid methodology that utilizes compartment modeling and the Discrete Element Method. The effectiveness of the methodology will be demonstrated by modeling particle mixing under the influence of an impeller in the continuous blender, which for usual modeling methods typically lead to extremely high computational costs. |
