(443h) Modelling Picking & Sticking on Pharmaceutical Tablets

Powder compaction is a common manufacturing process in many industries, for example those involving powder metallurgy, ceramics, pharmaceuticals, chemicals, food processing and nuclear fuel. The objective of any powder compaction process is to prepare compacts with desirable mechanical strength, minimal density gradients, shape within specified dimensional tolerances, and minimal flaws or cracks. To achieve this objective, the overall powder compaction process is frequently optimized to specify a formulation, i.e., a mixture of powders, and the tooling and process parameters that result in a desirable compaction performance. In this work, pharmaceutical tablet compaction is the application of interest.

In practice, pharmaceutical tablets are frequently formed with de-bossed surface features in order to uniquely identify the product, and they may be formed with a score or scores to facilitate tablet splitting. Debossing is one of the preferred methods for imprinting tablets since debossing can be easily incorporated into the manufacturing process by embossing the punches used to create the tablets.

Common problems faced when debossing include picking, ill-formed features, and stress concentrations that can lead to tablet fracture. Picking occurs when tablet material sticks to the embossed punch surface, resulting in missing regions of the features. Feature illegibility may result from poorly designed tooling or significant post-compaction elastic rebound of the tablet. Stress concentrations occur at concave debossed features such as interior corners and valleys, and if the radii of curvature are too small for a given tablet formulation and loading environment, then fracture of the compact can occur.

This work focuses on modeling the local elasto-plastic response of material in the vicinity of a debossed feature. Specifically, the degree of post-compaction material rebound is predicted along with the relative density (or solid fraction) field as changes are made to the debossed feature geometry and material formulation. The methodology presented here is the first step in developing a fully predictive model for investigating the manufacture of debossed features.