(544c) Predicting Hold-up Weight Changes in a Continuous Powder Mixer During Transitory Regimes

The mixing of solids is a common operation in several industrial applications like foodstuffs, cosmetics, ceramics, plastics, and drugs, etc?.The most prevalent process configuration used in the industry today is still the batch process. However, in these recent years, the continuous mixing process started to receive an increasing interest from various research teams in the world. In the past, we have contributed to the study of this process under steady-state conditions from both experimental and modelling points of view, for real cases, as well as for a model case. But the implementation of a continuous mixer in the industry needs the absolute control of the process during the transitory phases that are likely to occur: starting, emptying, feeder's feeding, accidental perturbation, etc. All these will have an effect on the hold-up weight in the mixer, and as a consequence on the mean residence time of the powder in the mixer. This will in turn affect the quality of the mixtures. A better understanding of the behaviour of the process, in particular in terms of hold-weight changes, during transitory regimes may therefore serve to the development of a strategy for process control.

This paper reports experimental and modelling results concerning the transitory operation of a continuous mixer. We investigate the effect of step variations in operating conditions, such as rotational speed of the stirrer or mass flow rates, on the bulk particle flow and transport in the continuous mixer, through prediction of the local hold-ups in the mixer. The apparatus used in this work is a pilot scale commercial mixer Gericke GCM500, for which a specific experimental protocol has been developed to determine the spatial distribution of the hold up weight inside the vessel during both steady and unsteady states.

A new stochastic approach based on Markov chains is developed to simulate the bulk powder dynamics inside the continuous mixer in transitory phases. A non-linear behaviour obtained during transitory regimes is modelled by a non homogenous chain, while, it is described by a homogeneous chain at steady state. Empirical correlations are developed to link the process variables and the matrix of transition probabilities P, which is the principal operator of a Markov chain. The application of this model allows the prediction of the evolution of the hold-up weight, the distribution of particles mass in the mixer and the intermediates flow rates inside the mixer, as well as that of the outflow rates during the continuous process, in both transitory and steady state regimes. The comparison of model results with experimental data not used in estimating parameters contributes to validating the viability of this model for a wide range of operating conditions, and constitutes a first approach to process control.