The modelling challenges of die compaction

Anton V. Kulchitsky
Research Associate Professor, University of Alaska, University of Alaska, Fairbanks, AK 75910, USA, CTO Coupi Inc.,
Jerome B. Johnson
Research Professor, University of Alaska, University of Alaska, Fairbanks, AK 75910, USA, CEO Coupi Inc., Fairbanks, AK, 99709,
Sergii Kutnii
Research Assistant, Coupi Inc., Kiev, Ukraine,
Gleb Velikhovskiy
Research Assistant, Coupi Inc., Kiev, Ukraine,

Modeling the die compaction of pharmaceutical powders into tablets is challenging due to the multiple particle types with different physical, chemical and geometrical properties and the variety of mechanisms involved in compacting powders. Pharmaceutical formulations contain an active ingredient, several excipients and a lubricant each with different grain size and shape distributions, cohesion and adhesion values, mechanical properties, chemical properties, and fracture response. Compaction proceeds through powder grain rearrangement, local elastic deformation, local and nonlocal plastic deformation, particle fracture, the formation of chemical and sintered bonds, pore filling, and air expulsion from, and entrapment in, tablet pore space. No model currently exists that can simulate the important physical processes that occur during compaction. Empirical models suffer from their limited applicability over a range of powder mixtures and compaction methods. Consequently, recent efforts attempt to describe compaction processes using numerical models of the micro-mechanical interaction of particles that include the discrete element method (DEM), the multi-particle finite element method (MPFEM), and combined FEM/DEM. At present, no single model approach is capable of fully simulating compaction. DEM models can simulate particle shape and size distribution, local elastic and plastic deformation, interparticle bond formation and fracture, MPFEM is used to simulate particle size and shape distribution and elastic and plastic deformation for local and nonlocal contacts with different particle mechanical properties, and FEM/DEM utilizes results from MPFEM simulations to derive algorithms for DEM models to incorporate local and nonlocal elastic and plastic deformation effects. Most modeling approaches use mono-sized spherical particles and no complete models of air flow or entrapment in tablet pores exists. DEM is computationally intensive, however, MPFEM is even more so. With the exception of air flow, many of the facets of compaction are contained in the various DEM, MPDEM, and FEM/DEM models. The challenge is to integrate these different facets into a single 3 dimensional model of compaction.

We demonstrate a modified DEM approach that captures many of the mechanisms of compaction using Coupi Polyphysica. We use special “truss” cluster particles built from hard spheres connected in a stable configuration with elastic rods resistant to tension, bending, and torsion deformations. We show that large ensembles of such cluster-particles reproduce the behavior of compressible powders and demonstrate non-local deformation under large confining stress. The powder can be compressed with the apparent solid fraction above 0.85 value which exceeds standard DEM approaches and match the observations. The model is less computationally demanding than an MPFEM approach and allows to simulate enough particles to represent powders. Although, the model does not reproduce the stress-strain in a single particle accurately, it reproduces the observed bulk behavior. We also suggest methods to integrate additional mechanism to allow simulation of the important features of powder compaction such as plastic deformations using these truss cluster DEM approach.


References:

Gonzalez, M., A. M. Cuitiño, 2012. A nonlocal contact formulation for confined granular systems. J. of the Mechanics and Physics of Solids 60: 333-350.
Gonzalez, M. 2019. Generalized loading-unloading contact laws for elasto-plastic spheres with bonding strength. J. of the mechanics and physics of solids 122: 633-656.
Haddad, H., Leclerc, W., Guessasma, M., Pélegris, C., Ferguen, N., & Bellenger, E. (2015). Application of DEM to predict the elastic behavior of particulate composite materials. Granular Matter, 17(4), 459-473.
Harthong,  B.,_Jérier, J.-F., Dorémus, P.,_Imbault, D., Donzé, F.-V., 2009. Modeling of high-density compaction of granular materials using the Discrete Element Method. International Journal of Solids and Structures 46: 3357–3364
Johnson, J. B. (1998). A preliminary numerical investigation of the micromechanics of snow compaction. Annals of Glaciology 26: 51-54.
Johnson, J. B. and M. A. Hopkins (2005). Identifying microstructural deformation mechanisms in snow using discrete element modeling. Journal of Glacionolgy 51: 432-442.
Leuenberger, H. & Rohera, B. D. 1986. Fundamentals of powder compression. I. The compactibility and compressibility of pharmaceutical powders. Pharm. Res. 3, 12–22 (1986).
Popescu, Ileana & Vidu, Ruxandra. (2018). Compaction Behaviour Modelling of Metal-Ceramic Powder Mixtures. A Review. Scientific Bulletin of Valahia University – Materials and Mechanics. 16. 28-37. 10.1515/bsmm-2018-0006.


Anton V. Kulchitsky

Dr. Anton Kulchitsky has degrees in Applied Mathematics (Masters) and Mechanics (Ph.D.) from Moscow State University. He has over 20 years of applied research experience, including 10 years in discrete element method development. He co-developed, with Jerome B. Johnson, the Polyphysica particle dynamics model. Dr. Kulchitsky has led research projects related to sea-ice drift modeling, NASA’s asteroid retrieval mission, and water migration in lunar regolith. He is a Co-I on the CAESAR comet sample return mission proposal, which seeks to return a volatile gas rich sample from comet 67P. He has also contributed to developing analytical and numerical methods to solve different practical problems, including intensive droplet evaporation, ionospheric advection, and ultra-low frequency electromagnetic wave propagation through the Earth’s crust. Currently Anton Kulchitsky is CTO of Coupi, Inc., where he leads the software development team. He is also a research associate professor at the University of Alaska Fairbanks.