Education - Lecture/Discussion - WPI Only
Wednesday, January 27, 2010
1:00 PM-2:00 PM
Higgins Laboratories
HL116
Professor Themis Matsoukas
Department of Chemical Engineering
Penn State University
Abstract : Particulate materials account for about 60% of chemical manufacturing and even among processes that do not produce particles, approximately one fifth of them involve dry powders or suspensions as ingredients at some intermediate processing step. Granulation is a unit operation of widespread use, especially in drug manufacturing. It is a process in which solid particles, typically an active pharmaceutical ingredient (API) and an inert excipient, are mixed together with a liquid binder solution in a fluidized bed. The binder acts as glue and causes particles to aggregate granules of homogeneous composition and desired size are obtained. In its core, granulation involves binary collisions between particles of different composition to form a larger particle in which components are partially mixed. We refer to this process as aggregative mixing to indicate the simultaneous evolution of size and composition by binary aggregation. Our goal is to formulate theory, models and numerical algorithms to predict the compositional distribution of the granules and its evolution in time. I will present a theory aggregative mixing of “ideal” components, i.e., systems in which the rate of aggregation is independent of the composition of the aggregating particles. The compositional distribution in this case is obtained in closed form and the state of mixing obeys a universal scaling that does not depend on the details of aggregation. Non-ideal interactions between components may lead to either improved mixing, or referential segregation, depending on the physical properties of the components.
I will discuss our Monte Carlo methodology which avoids the pitfalls of classical algorithms for the balance equations (PBE) and I will discuss comparisons with experiment. In the last part of the talk I will take a stochastic view of binary aggregation to show how tools borrowed from equilibrium statistical mechanics can provide insights into the character of irreversible processes involving macroscopic particles.
Suggested Audiences: College
E-mail:
mdemetri@wpi.edu
Phone: 508-831-5459
Last Modified: January 12, 2010 at 9:03 AM
Powered by the Social Web - Bringing people together through Events, Places, & Common Interests