Date of Award

Fall 1-31-2017

Document Type


Degree Name

Master of Science in Biopharmaceutical Engineering - (M.S.)


Chemical, Biological and Pharmaceutical Engineering

First Advisor

Piero M. Armenante

Second Advisor

Laurent Simon

Third Advisor

Michael A. Hanson


Knowledge of the minimum agitation speed, Njs, required to suspend finely divided solids in vessels stirred by an impeller is a critical parameter to properly operate industrial tanks in a large number of industrial operations. The most common experimental approach to measure Njs is that of Zwietering’s (Chem. Eng. Sci., 1958, 8, 244-253), consisting of visually inspecting the tank bottom and visually determining the impeller agitation speed at which the solids are observed to rest on the tank bottom for no more than 1-2 seconds before being swept away. This method is quite reliable, but a method not relying on the operator would clearly be preferred. Therefore, the main objective of this work was to develop a simple, repeatable, and observer-independent method to determine Njs for a variety of agitation systems and agitation conditions in tanks with a flat-bottom, the most common tank shape for which Njs has been obtained in previous studies. Flat bottom reactors are also among the most widely used reactors for research and development studies in the pharmaceutical and biopharmaceutical industry.

In this study, Njs was experimentally determined by using a newly developed method in which the area covered by the unsuspended solids still at the bottom of the tank was measured at increasing agitation speeds, starting at low speeds (below Njs). To do so, images of the tank bottom were captured in .jpg format by a digital camera. Each image was processed with the appropriate software (Image J) to quantitatively determine the area still covered by solids at that speed. Increasing the agitation speed increased the amount of solids being suspended, resulting in a decrease in the area covered by solids at the bottom of the tank. Plots of the area covered by the solids vs. the corresponding agitation speed resulted in a linear function, which when extrapolated to A→0 yielded the expected value of Njs (named Njs-A). The values so obtained for Njs-A were then compared to the Njs value determined visually (Njs-vis).

This approach was tested for a number of mixing systems with different impeller types (disk turbines, flat-blade turbine, pitched-baled turbines, hydrofoil impeller), impeller off-bottom clearances, impeller sizes, and tanks sizes. Preliminary tests were conducted using the same method but in hemispherical tanks, where the solids at the bottom of the tank a formed an approximate circular pattern. As compared to hemispherical bottomed tank, the solid particles in flat-bottomed tanks were dispersed all over the bottom of the tank. Therefore, the sum of all the individual areas covered by the solid particles deposited at the bottom was quantitatively determined. The results for Njs-A obtained with the new method were compared with those obtained using the traditional Zwietering’s approach. In general, excellent agreement was found between Njs -A and Njs-vis. It can be concluded that this newly developed method constitutes a novel, reliable, and, especially, observer-independent method to experimentally determine Njs.

It is expected that this approach could be of particular relevance in a variety of mixing applications, including the suspension of micro-carrier particles, such as glass beads, on to which animal cells can become attached and used to produce various bio-products, and in the suspension of beads for production of proteins.