Date of Award

Summer 2001

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Chemical Engineering - (Ph.D.)

Department

Chemical Engineering, Chemistry and Environmental Science

First Advisor

Piero M. Armenante

Second Advisor

Basil Baltzis

Third Advisor

Dana E. Knox

Fourth Advisor

Robert Pfeffer

Fifth Advisor

San Kiang

Abstract

Crystallization is a widely practiced unit operation throughout the chemical and pharmaceutical process industries. Despite its widespread application, crystallization still suffers a disproportionate number of process difficulties due to the complexity of the steps involved in the process. In particular, the final characteristics of the product are strongly affected by the mixing conditions during the process. The development of a validated modeling approach would be highly valuable for the successful prediction of the crystal characteristics under differing mixing conditions.

In the present work, the single-feed semi-batch precipitation process of barium sulfate in a stirred tank was experimentally studied and numerically predicted using a micromixing model based on CFD. A commercial CFD package (FLUENT) was used to simulate the flow field and predict the energy dissipation rate distributions within the reactor. The precipitation zone originated from the feed point was tracked using a random walk model. Available correlations were used for the calculation of the nucleation and crystal growth rates. The mass transfer coefficient for the crystal growth was assumed to be dependent on both the average crystal size and the local energy dissipation rate. Finally, a micromixing model (E-Model) was incorporated to predict the effects on the final crystal size distribution of a number of operating and geometric variables as well as the effect of vessel scale.

An extensive number of barium sulfate precipitation experiments were conducted to determine the crystal size distribution and validate the proposed model. The effect of the process variables (such as volume ratio, mean initial concentration and stoichiometry ratio), operating conditions (including impeller speed, diameter and off-bottom clearance), and vessel scale on the crystal size distribution were experimentally determined and numerically predicted. In general, very good agreement between experimental data and model predictions was obtained. The model was typically able to capture all of the most important features of the precipitation process. The proposed approach has a significant potential for the prediction of the performance of crystallization processes in industrial applications.

Share

COinS