Dingeman (Danny) van der Haven*, Frederik Ørtoft#, Maria Mikoroni#, Andrew Megarry#, Kaisa Naelapää#, Ioannis Fragkopoulos#, James Elliot*
*University of Cambridge, United Kingdom
#Novo Nordisk A/S, Denmark
Computer simulations of powder compaction have become an increasingly popular tool to assist the development of pharmaceutical tablets. Particularly successful so far has been the finite element method (FEM) in combination with the density-dependent Drucker-Prager Cap (dDPC) model. However, the use of these simulations is limited by the time and experimental data needed to create a successful parametrisation for a single powder mixture. This limitation makes computational studies unfeasible for even a modest number of powder mixtures. In this work, an automated parametrisation workflow is presented that quickly generates dDPC model parameters and requires minimal user input. Furthermore, a new mixing methodology is proposed to predict the behaviour of a powder mixture based only on pure-component data of the constituent powders. Compaction experiments were performed on micro-crystalline cellulose, dibasic calcium phosphate dihydrate, pregelatinized corn starch, and 9 mixtures thereof. Simulations were able to reproduce compaction curves with an average error of 3.0% of the maximum compaction pressure. Moreover, the mixture methodology predicted mixtures with an average error of 4.8%.
Danny van der Haven is a PhD student in the Macromolecular Materials Lab supervised by Prof. James Elliott at the University of Cambridge. His current work is on computational simulations of pharmaceutical powder compaction and in close collaboration with Novo Nordisk. Before starting his PhD, he obtained a bachelor’s in Biomedical Engineering and a master’s in Chemical Engineering at Eindhoven University of Technology with a focus on physical chemistry of soft matter.
You must be logged in to post a comment.