(26d) Scale-up of Dry Granular Material Flow in a Bladed Mixer

Drying of granular materials is a common unit operation for a wide range of industries. In the pharmaceutical industry, drying of Active Pharmaceutical Ingredient (API) crystals is often carried out in an agitated filter dryer. During the drying process, wet particles are heated in an agitated mixer which consists of a jacketed cylindrical vessel with a rotating impeller which promotes convective drying. Incomplete understanding of the process, results in operating problems such as under or over drying, non-uniform drying, polymorphism changes, agglomeration, and particle breakage. Moreover, it is difficult to choose the operating conditions to achieve the desired product due to several phenomena including mass transfer, heat transfer, and changes in material properties happening simultaneously in an agitated dryer. In addition, it is challenging to scale up the process from the laboratory scale to the pilot scale to the manufacturing scale. It is of interest to study the agitated drying process in order to optimize operating conditions for a particular material as well as develop and understanding of factors that affect scale up. In this study, we simplify the problem and we don’t examine flow, mixing, heat transfer and mass transfer together but instead we focus on the granular material flow in the agitated mixer. Dry, monodisperse, spherical glass beads were used as model materials in the mixer and the torque on the rotating impeller was measured in order to characterize the material flow. The experiments were designed such that flow occurred in the slow, frictional flow regime or quasi-static regime and we observed that impeller torque was independent of the shear rate (rotation rate of the impeller). We found that the height of the powder bed had a great effect on impeller torque whereas the particle size had less effect on impeller torque. We also found that a portion of the powder bed could be replaced with metal weights (in order to keep the same hydrostatic pressure) without changing the impeller torque. Previous work has suggested that shear stresses can be related to normal stresses for flowing granular materials. By plotting the normal stresses versus the shear stresses in the bed, we were able to compute a bulk friction coefficient and cohesiveness of the material, which are characteristic properties of the material, For scale-up studies, the bulk friction coefficient and cohesiveness of the material were used to predict the impeller torque. We illustrated the prediction of impeller torque of the material by examining a bed of large height and observed good agreement between our predictions and measurements in the lab. We will discuss how this work serves as an efficient way to predict the granular material flow in a larger scale using data from the lab scale. We will also discuss the application of these results to scale up of agitated drying of pharmaceutical materials.