Abstract
In this paper, the compaction of lactose powder, a typical pharmaceutical excipient, is modelled using finite element methods (FEM) in which the powder is represented by an elastic-plastic continuum medium following Drucker-Prager Cap yield criteria. In a recent numerical and experimental study by the present authors [1], it was found that cone-shaped capping failure occurs during compaction of flat-faced round tablets, and that cone capping is associated with intensive shear band formation during the decompression stage. It is hence instructive to explore possible approaches that might alleviate the propensity for capping. Two approaches to alleviate capping were therefore investigated through finite element analysis: (i) altering the surfaces of the punches, i.e. to make convex tablets using the same material properties, and (ii) altering the material properties, i.e. changing the elasticity of the materials. It was found that capping still takes place even if the surface curvatures of the punches are altered. These predictions have been confirmed by physical experiments using a compaction simulator. The experiments have also demonstrated convincingly that the capping occurs during decompression. The second approach has been investigated in such a way that only the Young's modulus of the powder is changed to values twice and one-half that of lactose. Numerical results reveal that intensive shear bands are still developed during decompression even when the material properties are changed in this way. This implies that similar capping patterns are still possible for those materials. It is anticipated that the reason that some pharmaceutical excipients, such as microcrystalline cellulose (Avicel PH-102), do not cap is because of the high bonding strength of such materials (which can be generally characterised by tensile strength) [2].