(371f) Application of a Combination of Statistical Techniques and Mechanistic Modeling to the Mixing Process in Co-Rotating Twin Screw Extruders

This paper proposes and experimentally validates theoretical models for the mixing process in a plasticating co-rotating twin screw extruders using a combination of statistical techniques and mechanistic modelling. A predictive model for the global residence time distribution was developed using a continuous exponential distribution model by examining the local contributions. Parameters such as the mean residence time, mean time delay, degree of fill, local drag and volumetric flow rates were obtained using mechanistic modelling by mathematically modelling the flow dynamics in the extruder. Predictive models of the mean residence time and time delay were subsequently developed. The effect of input conditions such as feed rate and screw speed on the shape of the residence time distribution curve and the degree of mixing were also evaluated by analysing the variance in this distribution.

Increase in feed rate at constant screw speed and vice versa results in narrowing the shape of the residence time distribution curve, with the feed rate having a higher sensitivity. The mean residence time and time delay was observed to drop with an increase in the feed rate. With an increase in the screw speed at constant feed rate, a reduction in the mean residence time was observed at the expense of an increased time delay. This has been validated against the experimental results.