A stochastic technique for multidimensional granulation modeling

doi:10.1002/aic.10911

2006-06-26

PRE

Justin A. Gantt

Edward P. Gatzke *

DOI：10.1002/aic.10911

原文链接

原文求助

摘要

En 中文

Recent granulation modeling research has produced compelling evidence that simple one-dimensional (1-D) models will not suffice when describing the dynamics of particle growth. During the mixing process particles gradually become more saturated due to the loss of air in particles resulting from collisions with the surroundings. This process, called consolidation, influences not only granule pore saturation and binder layer thickness, but also particle coalescence properties. Due to the consolidation behavior observed, a simple size dependent model is typically not adequate. The solution of a multidimensional population balance equation (PBE) is performed using a multiscale solution technique. A microscale model which utilizes a discrete element method (DEM) simulation is used to determine granular flow characteristics and coalescence efficiency. Statistical analysis of data from interactions at the microscale is then used to produce a multidimensional coalescence kernel. Since analytical solutions to PBEs with more than one internal dimension typically do not exist, a constant number kinetic Monte-Carlo (MC) method is used to solve the PBE. This solution technique is used in a test case to model growth of a lactose/starch/HPC and water mixture. The simulated results corresponded well with experimental data. This technique, while in its infancy, proves to be a unique and effective method of bridging multiple scales of the granulation process. (c) 2006 American Institute of Chemical Engineers.

Keyword:

population balance

kinetic Monte-Carlo

discrete element method

coalescence kernel