Analyzing Homogeneity of Highly Viscous Polymer Suspensions in Change Can Mixers

Abstract

The mixing of highly viscous non-Newtonian suspensions is a critical process in various industrial applications. This computational fluid dynamics (CFD) study presents an in-depth analysis of non-isothermal mixing performance in change can mixers. The aim of the study was to identify parameters that significantly influence both distributive and dispersive mixing in these mixers, which are essential for optimizing industrial mixing processes. The study employed a numerical design of experiments (DOE) approach to identify the parameters that most significantly influence both distributive and dispersive mixing, as measured by the Kramer mixing index ((Formula presented.)) and the Ica Manas-Zloczower mixing index (Formula presented.). The investigated parameters included mixing time, number of arms, arm size ratio, revolutions per minute (RPM), z-axis rotation, z-axis movement, and initial and mixing temperatures. The methodology involved employing the bootstrap forest algorithm for predicting the mixing indices, achieving an (Formula presented.) of 0.949 for (Formula presented.) and an (Formula presented.) of 0.836 for (Formula presented.). The results indicate that the z-axis rotation has the greatest impact on both distributive and dispersive mixing. An increased number of arms negatively impacted (Formula presented.), but had a small positive effect on (Formula presented.). Surprisingly, in this study, neither the initial temperature of the material nor the mixing temperature significantly impacted the mixing performance. These findings highlight the relative importance of operational parameters over traditional temperature factors and provide a new perspective on mixing science.