Document Type
Dissertation
Date of Award
12-31-2021
Degree Name
Doctor of Philosophy in Chemical Engineering - (Ph.D.)
Department
Chemical and Materials Engineering
First Advisor
Piero M. Armenante
Second Advisor
Ecevit Atalay Bilgili
Third Advisor
S. Basuray
Fourth Advisor
Roman S. Voronov
Fifth Advisor
Kevin G. Reuter
Abstract
The hydrodynamics of process units used in pharmaceutical industry, i.e., the distribution of velocity profiles and shear stresses within the vessel, often controls the process performance, significantly affecting product quality. The hydrodynamics is strongly influenced by several factors such as the vessel shape, stirrer type, presence of internals and baffles, agitation intensity, and liquid properties. Most of these parameters can be varied by the process designers and the operators to achieve the desired process outcomes. The impact of the hydrodynamics on the process outcome can be profound in both large vessels (e.g., reactors) and smaller vessels (e.g., dissolution vessels), involving one or multiple phases. Once the geometry of the processing unit, its operation characteristics (e.g., agitation speed) and the fluid properties has been defined, the system develops its own hydrodynamic behavior and, thus, affects other relevant mixing parameters, such as power dissipation, turbulence intensity, blend time, just-suspended speed, and others.
Quantifying the hydrodynamics by itself can be extremely useful to understand how the process and variability of results are affected by changing different operational or geometric variables, especially so in pharmaceutical applications due to the stringent requirements typically imposed on pharmaceutical products. For these reasons, it is becoming increasingly evident that a more fundamental understanding of the hydrodynamics and mixing characteristics of such devices is essential to allow more practical designs and operational strategies.
Therefore, the overall objectives of this work are to quantify the hydrodynamic characteristics of different vessel configurations of pharmaceutical equipment stirred by various types of agitation systems through the use of experimental and computational approaches. Particle image velocimetry (PIV) is the primary experimental tool used in this work. In addition, to achieve the most effective and meaningful approach for studying flow fields, including scale-up parameters, a computational approach using Computational Fluid Dynamics (CFD) is employed and validated with experimental results. More specifically, in this work the hydrodynamics of the USP Dissolution Testing Apparatus 1 (basket type) is studied in significant detail, both experimentally and computationally, resulting in a full mapping of the velocity flow field for different fill levels, basket mesh sizes, and agitation speeds. In addition, the blend time in this system is experimental measured and computationally predicted.
The hydrodynamics of other larger systems used in API manufacturing is also studied in this work. Specifically, the flow field and the power dissipation characteristics of mechanically stirred, glass-lined vessels provided with a Retreat-Blade Impeller (RBI) and one of different types of single baffles is obtained since, despite their common use, little information is available on the power dissipated by the RBI in such systems. The turbulent Power Numbers, Po's, for an RBI in a scaled-down pharmaceutical type of vessel equipped with one of the different types of single baffles used in the pharmaceutical industry (beavertail, D-type, H-type, finger-type, fm-type baffles) are experimentally determined for different positions of the baffles and computationally obtained, together with the velocity distribution and strain rate. The experimental Po values are in close agreement with the computational predictions. A simple power-law correlation is established between Po and the dimensionless vertical cross-sectional area of the baffles, resulting in an equation with a good fit with the experimental data. In addition, a correlation between Po and the dimensionless baffle area exposed to the tangential flow generated is developed, which could have a very general applicability.
Finally, the determination of the minimum agitation speed, Nis, to achieve the just-suspended off-bottom solid suspension state in liquids in stirred vessels is an issue of significant importance in pharmaceutical processes. Here, Nis is computationally predicted for a stirred, fully baffled vessel provided with different axial or radial impellers using an LBM-based CFD model coupled with a novel method to extract Nis from the computational results. Accordingly, the number of solid particles in a very thin control volume near the bottom of the vessel is computationally predicted over a range of agitation speeds, N, to determine the mass fraction of suspended solids, Xm. A regression analysis based on the logistic equation is then applied to the Xm-N curves and a simple, derivative-based mathematical method is applied to predict Nis. The results obtained with this novel computational approach are found to be in very good agreement with experimental data.
In summary, this work is focused on the quantification of the hydrodynamics of a number of apparatuses, large and small, commonly used in the pharmaceutical industry, and the impact on other i phenomena of significant importance such as solids suspension or vortex formation. It is expected that the results of this work will be of significant relevance to engineers and scientists working with such systems and to fellow researchers studying them.
Recommended Citation
Sirasitthichoke, Chadakarn, "Hydrodynamic characterization of laboratory and production equipment for pharmaceutical testing and manufacturing using experimental and computational approaches" (2021). Dissertations. 1786.
https://digitalcommons.njit.edu/dissertations/1786
Included in
Chemical Engineering Commons, Fluid Dynamics Commons, Pharmacy and Pharmaceutical Sciences Commons
