ABSTRACT
The drying capacity of a continuous vibrated fluid bed dryer was studied using a DoE by varying microcrystalline cellulose content in the formulation, water amount in the twin-screw granulation, inlet air temperature, air flow rate and the acceleration of the horizontal fluid-bed. Temperature and humidity profiles were measured along the dryer using wireless sensors. For the parameter space explored in this study, acceleration was the most influential process parameter of the dryer regarding the resulting granule moisture content. An empirical model was developed that allowed for fast and accurate moisture content prediction that could be incorporated into an enhanced control strategy. In addition, a mechanistic model was formulated that allow for prediction of temperature and moisture profiles, and most importantly the moisture content of the granules inside the dryer. The mechanistic model can be integrated to other unit operation models to provide overall understanding of an integrated continuous process line. The mechanistic model also makes it possible to define the equipment design requirements (e.g., length of the dryer) to meet the specific needs in terms of drying capacity, temperature and moisture profile.
ABSTRACT
The tablet pan coating process involves various types of transverse tablet bed motions, ranging from rolling to cascading. To preserve satisfactory results in terms of coating quality after scale-up, understanding the dynamics of pan coating process should be achieved. The aim of this study was to establish a methodology of estimating translational surface velocities of the tablets in a pan coater and to assess their dependence on the drum's filling degree, the pan speed, the presence of baffles and the selected tablet properties in a dry bed system and during coating while varying the drum's filling degree and the pan speed. Experiments were conducted on the laboratory scale and on the pilot scale in side-vented pan coaters. Surface movement of biconvex two-layer tablets was assessed before, during and after the process of active coating. In order to determine the tablets' surface flow velocities, a high-speed video of the tablet surface flow was recorded via a borescope inserted into the coating drum and analysed via a cross-correlation algorithm. The obtained tablet velocity data were arranged in a linear fashion as a function of the coating drum's radius and frequency. Velocity data obtained during coating were close to those of dry tablets after coating. The filling degree had little influence on the tablet velocity profile in a coating drum with baffles but clearly affected it in a coating drum without baffles. In most but not all cases, tablets with a lower static angle of repose had tablet velocity profiles with lower slopes than tablets with higher inter-tablet friction. This particular tablet velocity response can be explained by case specific values of tablet bed's dynamic angle of repose.