Abstract
As one of critical process steps during pharmaceutical tabletting, rotary die filling is still not well understood. To address this issue, a model rotary die filling system with a paddle feeder was developed to closely mimic the industrial process. Using this model system, the performance of various pharmaceutical powders at different turret and paddle speeds was evaluated, and the dependence of fill variation on process conditions and material properties was examined. A comprehensive dataset was created and reported here to show the effects of material and process parameters on the die filling performance and the filling consistency. It is believed that the data can also be used for data-driven process modelling and for developing robust machine learning models for pharmaceutical manufacturing.