(432h) Segregation of Competing Mechanisms

Segregation is a mechanistically driven event. That means it has several causes. It is important that formulators understand these causes so as to know how or what to change in the formulation to facilitate smart product design. Without this knowledge, formulators are left with a trial-and-error approach, and often get lost in conducting a vast amount of testing and characterization using multi-factor experiments. Fines sift through a matrix of coarse particles, air currents present during filling carry fines to different parts of the process, particles with different frictional properties or shape roll or slide down piles at different rates, particles with different coefficients of restitution bounce down the pile with different trajectories, or entrained gas within the particle bed can be released upward and carry fines to the top surface. Each segregation mechanism induces the separation of particles during processing. Each segregation mechanism also induces a pattern of segregation based on key particle scale properties that can be easily characterized. One could create a vast computational method using DEM and FEM models and then let the models calculate to generate the overall segregation pattern and magnitude. But that brute force approach still does not give the formulator any means of knowing how to change the formulation to prevent or minimize segregation. Nor does the brute force method provide process engineers with any means of knowing what to change in the process to mitigate segregation. A superposition principle may help. Each segregation cause (mechanism) will induce a unique pattern where the magnitude of segregation of the various components is dependent on the key particle scale properties inducing the segregation (i.e. particle size, surface friction, coefficient of restitution, or aerodynamic radius). If the overall segregation pattern and magnitude are known, and the segregation pattern for each mechanism is understood, then that information can be combined to yield a means of determining the root causes of segregation on a component-by-component basis – thereby giving formulators and process engineers tools to estimate how to change the process or the product to mitigate segregation. This paper looks at an ideal system with three active mechanisms (sifting, repose, and rebound) and presents a method of determining the root causes of segregation for each component in the system.