GOALI: A Multiscale Modeling Framework for Predicting Granule Property Evolution in Mixer Granulators

Purdue University, West Lafayette IN

Investigators

Abstract

The goal of the project is to develop fundamentally-based, multiscale modeling tools that will aid in the understanding and scaling of granulation processes, an important step in the manufacture of new products and delivery forms in detergents, consumer goods, pharmaceuticals, agricultural chemicals and specialty materials. The proposed work focuses on developing a validated and coupled discrete element method (DEM) / population balance (PB) computational model for predicting the evolution of granule size and liquid content in a dual axis paddle granulator. Specific objectives of this project are to: 1. Develop a DEM model that can give predictions of particle velocity fields and residence times in the spray zone. The residence time predictions will be validated against high speed video residence time distributions and Positron Emission Particle Tracking (PEPT) data for an identical system. 2. Develop a multidimensional population balance model to predict the evolution of granule size and liquid content, and validate this model with laboratory scale measurements of multidimensional granule distributions; 3. Link the DEM and PB models in a serially integrated multiscale framework and use the combined model to predict performance in a lab scale granulator and the Procter and Gamble granulation pilot plant. The work proposed here primarily targets layered growth of seed granules. This work is part of a larger effort to develop a fully generalized multiscale model including nucleation, densification, coalescence, and breakage of fine powders. Intellectual Merit: This proposal is the first serious attempt to develop a multiscale model for design of granulation processes and validate the model with multidimensional distribution data at laboratory and pilot scales. This is a challenging class of problems in particulate processing because there is very strong two way coupling between the dense particulate flows and the granule growth processes. The scientific originality and significance of the research includes: -Special care in understanding the sensitivity of the DEM model predictions of macroscale phenomena to assumed mechanical properties of the granules; -Incorporating for the first time the correct physics of the spray zone into layering kinetic expressions for the population balance; -A pioneering study of DEM and PB model integration using a two way serially integrated multiscale framework; and -Unique and substantial experimental validation and application of the design models to pilot scale operation. Broader Impact: The multiscale modeling approach will be generalizable to all wet granulation systems. The models will dramatically transform the way these processes are designed. Currently, most granulation processes are designed by trial and error resulting in significant loss of materials, not to mention time, potential profits, and product effectiveness. Giving engineers the tools that allow them to more accurately design and control granulation operations will have impact over all industries that manufacture consumer and food products, pharmaceuticals, and agricultural chemicals. Successful completion of the proposed objectives will move wet granulation from an art, to a predictable and quantifiable science. This project will have a substantial educational impact at many levels: learning materials for K-12 demonstrations (developed by undergraduates in the Engineering Projects In Community Service program at Purdue), core undergraduate curricula, and industry short courses. These educational tools, as well simplified modeling tools, will be shared with the international community through the unique cyberspace community on www.pharmahub.org. The project also provides excellent training for a graduate student including an industry internship.