(239d) Development of an Empirical Model to Predict the Mean Residence Time in a Tablet Press Feeder

The die filling process is a continuous operation that is crucial to comply with the specifications and quality attributes of a pharmaceutical tablet, is usually performed using a feed frame. An experimental investigation was carried out to study the pharmaceutical powder behavior inside the feed frame and to develop an empirical model to predict the mean residence time in a tablet press feeder, having the feed frame paddle wheel speed, die disc speed and the properties of the materials as model parameters. Pulse injection method and sampling at feed frame output was used to study the residence time distribution (RTD). The Taylor dispersion model was used to fit the experimental data. Based on the fit, mean residence time (MRT) and mean centered variance (MCV) were calculated. A good performance and reproducibility of experimental method was found based on relative standard deviation of the mean residence time for one experiment. The results showed that paddle wheel speed, die disc speed and the properties of the materials affect significantly the mean residence time. The empirical model shows a linear relationship between the mean residence time and the paddle wheel speed, as well as the mean residence time and the die disc speed. However, the model indicates a non-linear relationship with the cohesion parameter, the selected property of the materials. Six independent experiments in which operating conditions and the cohesion parameter were controlled and varied were performed to validate the empirical model. Overall, the predicted mean residence time has a high correlation with the observed experimental results, finding error percentages less than 3.2% for the validations performed.