Date of Award

Spring 2001

Document Type

Dissertation

Degree Name

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

Department

Chemical Engineering, Chemistry and Environmental Science

First Advisor

Samir S. Sofer

Second Advisor

Ching-Rong Huang

Third Advisor

Dana E. Knox

Fourth Advisor

David S. Kristol

Fifth Advisor

Arthur B. Ritter

Abstract

This dissertation investigated the centrifugal, batch separation of whole blood into subpopulations of red blood cells (erythrocytes) and white blood cells (leukocytes). Separations took place in a custom-built centrifuge (using a seal-less, anti-twisting feed/withdrawal system) containing, a 25-ml capacity separation chamber. The blood separation chamber had dart-shaped geometry in the radial plane and a constant depth in the axial direction. Separation experiments were performed on whole bovine blood at varying hematocrit, centrifuge speed, and batch duration. A small, companion study of whole human blood separation runs also were conducted; they concentrated on batch duration effect and achieved superior separations.

A new graphical technique-generating accumulated cell-fraction separation graphs and measuring separation quality-was devised to display experimental separation runs. Results were presented for both bovine blood and human blood. An interval, observable between the accumulated cell-fraction curves of red blood cells and white blood cells, was measured and used to quantify the maximum extent of separation, allowing for determination of good and bad separations. This measured value was labeled separation quality (SQ). Measurements of SQ for bovine blood separation runs of various duration showed that batch duration had a strong correlation to separation quality. The set of human blood separation runs demonstrated that SQ values may be used as a means to locate optimal operating parameter values. An optimum was bounded for the human blood data set.

A one-dimensional volume-diffusion model has been derived for the equations of change of fluid mechanics. The volume-diffusion model extended the original work of Bird, Curtiss, and Hirshfelder in the area of molecular diffusion to application on particulate systems where volume diffusion was the predominant driving force. This model described the binary system of red blood cells (erythrocytes) and plasma. Expressions for volume flux with respect to stationary coordinates, including contributions via ordinary diffusion and pressure diffusion, were derived from the molecular flux expressions for the corresponding diffusion contributions. Due to its high degree of complexity, the model's system of partial differential equations could not be solved using a collocation finite element solver. The model was intractable.

Share

COinS