(465d) Modeling Density and Temperature Distributions in Powder Die Compacts

Tableting is one of the most important unit operations in the pharmaceutical industry because more than 90% of drugs are delivered in solid dosage forms. The physical and mechanical properties of tablets such as strength, density, and temperature are directly related to this process. The ability to predict through finite element (FE) modeling the physical and mechanical properties is essential towards maintaining a desired design space. For example, FE predictions of density distributions and temperature, which can alter dissolution rates and the polymorphic state of the active ingredient, respectively, could result in a change outside the design space. Therefore FE simulations are an important aspect of establishing a design space for support of quality by design for pharmaceutical processing. Previous FE modeling attempts were conducted in 2D (axisymmetry) but this current research explores the full 3D nature of the problem with the model material Avicel PH102.

The fully coupled thermomechanical FE software package ABAQUS provided a means to implement the Drucker-Prager Cap model which describes the evolution of the yield surface with varying relative densities. In addition, Young's Modulus and Poisson's ratio were varied with respect to RD to describe the elastic properties of the powder during unloading. A single compaction, unloading, and ejection step were analyzed for a capsule shaped tablet. Density variation and the temperature history of every point within the tablet is predicted. Furthermore, microcomputed tomography density measurements and infrared camera temperature measurements of post compacted tablets from a fully instrumented compaction simulator were used to validate the model.

The FE simulation agrees with the experimental force-displacement profile of the punches throughout compaction. The density predictions of the model are further validated with microcomputed tomography measured relative density values as shown through several internal cross sections. In addition, the model is capable of predicting surface temperatures of the post ejected tablets as supported by infrared camera temperature measurements.

The model is capable of predicting the full 3D nature of density and temperature distributions and further validates the implementation of the Drucker-Prager Cap model for pharmaceutical tableting. Moreover, the ability of FE to predict such properties presents the opportunity to further optimize tablet design, tool design, lettering placement, sticking, and overall process development.