<p>Subspace Identification Methods (SIMs) are a class of new identification methods that have drawn considerable interest in recent years. The key idea of these methods is to estimate the process states or the extended observability matrix directly from the process input and output data. The best-known SIMs are Canonical Variate Analysis (CVA), Numerical Subspace State-Space System Identification (N4SID) and Multivariable Output Error State space (MOESP). This thesis focuses on both fundamental research and application study of SIMs. The first part of the fundamental research involves the analysis of SIM algorithms from a statistical estimation viewpoint. For this purpose, a multi-step state-space model is set up first to reveal the relationships between the process states and the process data sets. Based on this model, SIM algorithms are analyzed to reveal their basic principles and bias issues. Several new SIM algorithms are proposed and shown to have similar performance as the existing algorithms. Relationships between SIMs and Latent Variable Methods (LVMs) for identification are then explored. It is shown that N4SID can be derived from Reduced-Rank Analysis (RRA) just as CVA is developed based on Canonical Correlation Analysis (CCA). Insights from this relationship lead to a variety of approaches to improve the performance of N4SID. The similarities and differences between SIMs and LVMs are investigated, with emphases on their causality, data collection and applications. For estimating the states, CCA and RRA are shown to be more efficient than Principal Component Analysis (PCA) and Partial Least Squares (PLS). A general statistical framework is proposed to unify SIM algorithms. The framework breaks all SIMs down into three common steps: (1) use of a linear regression method to estimate the predictable subspace, (2) use of a latent variable method to estimate a minimal set of the state variables, and (3) then fitting the estimated process states to the state-space model. Combining the approaches in the first two steps leads to a whole set of new SIM algorithms. Simulation studies show that these new SIM algorithms have similar performance as the existing SIMs. This framework reveals the nature of the computation steps in SIM algorithms and the fundamental ideas behind SIM algorithms. It also discloses the relationships among different SIM algorithms. The applicability of SIMs for closed-loop data is investigated. The original N4SID algorithm and the CVA algorithm based on regressing out the effects of future inputs are shown to give biased results. In general, whether a subspace identification algorithm is applicable for closed-loop data depends on how the effects of future inputs are treated in estimating the predictable subspace (step one of the proposed framework). Based on this analysis, several new N4SID and CVA algorithms are proposed for closed-loop data. Practical issues arising from applications of SIMs are also discussed. SIMs are shown to be able to handle the delays, common dynamics, non-stationary and co-integrating disturbances in the process. The advantages, as well as solutions for possible problems, are also presented. Some general guidelines are provided for applications of SIMs.</p> / Doctor of Philosophy (PhD)

<p>The synthesis and characterization of various thiolated polyethylene oxide (PEO) polymers constitute a significant part of the work, from linear PEOs with different molecular weights, and different end groups, to star-type, multiarmed PEOs, to linear PEOs in which one end bears a thiol group and the other a bioactive group. Different reaction protocols were used for the synthesis of the different PEOs. Terminal thiol groups were attached to PEO by reaction with mercaptoacetic acid or cystamine. Thiolated PEOs were chemisorbed to gold films deposited on polyurethane or silicon wafer substrates. Combinations of PEO and chain-end functionalized PEO were used to passivate the gold films and impart specific bioactivity respectively. The interactions of the modified surfaces with blood and blood proteins were investigated using radiolabeling, electrophoresis and immunoblotting methods. In measuring protein interactions with gold-based surfaces using radioiodine-labeling methods, it was found that the radioactive iodide ions (125 I- ) in the protein solution became bound to the surface along with the protein. Data on the adsorption of fibrinogen and albumin from buffer and plasma onto surfaces modified with different PEOs showed that adsorption decreased with increasing PEO molecular weight, irrespective of the group on the "outer" end of the PEO (hydroxyl or methoxyl). Surfaces were also prepared using PEO to which heparin was conjugated. Unfractionated heparin and low molecular weight heparin conjugated PEO were used to modify both gold and polyurethane surfaces. Both the passivation effect of PEO and the anticoagulant effect of heparin were found to be evident for these surfaces. (Abstract shortened by UMI.)</p> / Doctor of Philosophy (PhD)

<p>Polymers are routinely used in paper making to enhance the strength of paper. The strength enhancing effect of polymers on paper is generally attributed to the increase in the fiber-fiber bond strength. This thesis focuses on the links between the structures and interactions of adsorbed polymers and the resulting strength of fiber-fiber bonds. Dextran was chosen as the model of strength enhancing polymers in this work. The molecular weight, charge density, and hydrophobicity of the dextrans were varied. Mixtures of dextran and hydrophobically modified dextrans were employed to investigate the role of adsorbed polymer compatibility in fiber-fiber bonding. Cationic dextran with different charge densities (0.161-0.488 meq/g) was prepared. Hydrophobically modified dextran was made by the reaction of dextran with straight chain saturated C3, C4 and C6 fatty acids. The incompatibility of dextran and hydrophobic dextran was evaluated by means of the biphase formation of an aqueous mixture of the two polymers. The adsorption behavior of dextrans on pulp fiber was measured. It was observed that the maximum amount of adsorption increased with a decrease in the molecular weight and charge density of the dextran. The strength of paper was evaluated using tensile strength and Scott bond tests. The strength enhancing effect of dextran was given as a function of molecular weight, charge density, hydrophobicity, and incompatibility with hydrophobic dextran. When fiber surface was saturated with dextran, it was found that the tensile strength of paper did not depend on the molecular weight of the dextran in the range investigated (77,000 to 2000,000). The charge density of the dextran affected the strength of paper by changing the maximum amount of dextran adsorbed on fibers. Introducing hydrophobic groups into dextran decreased the strength enhancing ability of dextran. The strength enhancing mechanism of dextran was proposed to be the formation of dextran intermolecular bonds between fibers. For the first time, the role of polymer incompatibility in determining the strength of paper was demonstrated in this work. It was found that the bond strength between fibers coated with two incompatible polymers was weaker than that obtained from the single polymer coated fibers. The Page equation, along with the model developed in this work, was applied to calculate the bond strength for each type of bonding.</p> / Doctor of Philosophy (PhD)

<p>The present work is concerned with the numerical simulation of air-cooling in the film blowing process. In many film blowing production lines, film cooling is the limiting stage in achieving higher productivity. At the same time, the cooling air stream affects the stability and the shaping of the molten bubble, as well as the morphology of the final film product. The cooling process ultimately affects both production rate and final film properties. The aerodynamics of the external cooling air and the effect of air-ring design and operational setup on the cooling efficiency were examined, using turbulent airflow simulations based on the k-[varepsilon] theory. Both single and dual air rings designs were studied. The results suggest that the air-ring airflow is dominated by the Venturi and Coanda effects and that the airflow patterns are very sensitive to minor air-ring design modifications. The variation of heat transfer coefficients along the bubble surface has been examined in detail. Additional numerical simulations were performed to evaluate the performance of a typical internal bubble cooling (IBC) configuration. Some of the limitations of typical IBC implementations were identified and attempts were made to improve the designs using the numerical simulation as an optimization tool. The results suggest that numerical simulations can be used to gain valuable insight on the IBC operation. Numerical simulation can be helpful in reducing the number of trial and error steps during the design and implementation of IBC systems. The development of sizeable temperature gradients inside the melt (in the film thickness direction) was investigated using numerical simulation. Typically, large air-cooling rates at the film surface combined with the low thermal conductivity of polymers lead to significant temperature differences between the internal and external film surfaces. The result indicate that the temperature differences in the film thickness direction may be very large and, therefore, important from a design and modeling perspective. A new methodology to calculate the film stresses is proposed, in which the temperature variations in the thickness direction and their effects on melt rheology during the blowing are taken into account. The results indicate that the film reaches the crystallization temperature having stress differences in the thickness direction. At high production rates the stress differences became more pronounced. Since the film stresses are directly related with the crystallization kinetics, it is conjectured that the final film morphology and properties are affected by the predicted stress differences in the thickness direction.</p> / Doctor of Philosophy (PhD)

<p>Most sources of drinking water supply contain dilute concentrations of natural and synthetic organic substances which are not readily degraded by micro-organisms. Many of these compounds are known or suspected carcinogens, and/or may act as precursors for the formation of chlorinated hydrocarbons during the chlorination disinfection processes practiced in water treatment plants. The removal of these substances prior to chlorination is therefore desirable.</p> <p>The objective of this study was to test the claim that the oxidation of these organic substances by ozone results in the formation of more readily biodegradable compounds. A 25 ℓ laboratory ozone contactor was designed and built to simulate the ozonation process in a typical water treatment plant. The apparatus was used to ozonate samples of surface water and dilute synthetic aqueous solutions at applied doses ranging from 20 mg/ℓ to 40 mg/x. The biodegradability of ozonated and unozonated 10 ℓ samples was then evaluated in electrolytic respirometers, based on the changes in soluble Total Organic Carbon, Chemical Oxygen Demand, UV absorbance, and Biochemical Oxygen Uptake measurement.</p> <p>The ozonation system successfully simulated full scale operating conditions, achieving greater than 90% ozone utilization. Total Organic Carbon removal was described by a first order system with reaction rate constants ranging from 0.0041 min‾¹ to 0.023 min‾¹.</p> <p>Ozonation was found to improve the biodegradability of refractive organic substances in the three water sources examined. However the results show considerable variation depending on the particular source ozonated. Generally higher applied ozone doses improved biodegradability. Total Organic Carbon removal for the combined ozonation/biodegradation process ranged from 10% to 70% depending on the specific ozone dose and water source.</p> <p>For oxygen uptakes less than 2 mg/ℓ the respirometry equipment provided only qualitative evidence of bacterial activity. It is recommended that further studies should evaluate the use of an electrolytic respirometer which is independent of atmospheric pressure fluctuations.</p> / Master of Engineering (ME)

<p>This thesis focuses on developing empirical methodologies for improving process operation and product quality, in four important chemical engineering problems, using multivariate projection methods, such as Principal Component Analysis (PCA) and Projection to Latent Structures (PLS). The four problems addressed in this work are concerned with (i) improving and optimizing the trajectories of manipulated variables in batch processes; (ii) improving the identification of non-parsimonious dynamic process models using the Jackknife and the Bootstrap methods; (iii) developing meaningful specification regions for raw materials entering a consumer's plant and, (iv) improving transition policies in start-ups, re-starts and grade changeovers that are routinely performed in multi-product plants. The first problem addresses the situation where one desires to gain understanding of how and when, during the course of a batch, manipulated process variables have a significant effect on product quality. This amounts to estimating the sensitivity of product quality to manipulated process variables at various degrees of completion in a batch process. This information can be used in process development and in the optimization of already existing processes. The proposed approach involves adding designed experiments to batch policies currently used and then analyzing the resulting data bases using multi-way multi-block PLS. A new pathway PLS algorithm was developed for incorporating intermediate quality measurements collected during the course of each batch. In the second problem, the identification of non-parsimonious dynamic process models is improved through a more judicious selection of the meta parameter in regularization methods (ridge regression) and latent variable methods. These methods are often used to overcome ill-conditioning frequently encountered in the identification of such over-parameterized models. A new criterion for selecting the meta parameter (ridge parameter in regularization methods and the number of components in latent variable methods) is proposed, based on Jackknife and Bootstrap statistics. It is shown that this criterion outperforms the use of cross-validation (default criterion) and leads to the identification of models that are closer to the true process behavior. Developing an approach for defining multivariate specifications on incoming raw materials was important because there is a void in the quality control literature in this area. Specifications are usually defined in a univariate manner, based on past and often subjective experience. This work provides a sound, data-based approach for developing truly multivariate specification regions in a variety of industrial situations. The approach uses PLS methods to analyze historical data on the incoming raw materials, on the consumer's plant, and on the consumer's end product to define multivariate specification regions for the incoming raw material properties. The last problem consists of improving the performance of process transitions (start-ups, re-starts and grade changeovers), using historical process data. In particular, this work addresses two questions: (i) how to improve transition policies to minimize transition time and amount of off-grade materials, while ensuring safe operating conditions and (ii) how to ensure that at the end of a transition, steady-state process conditions are such that good quality products are obtained, and are consistent with past periods of production.</p> / Doctor of Philosophy (PhD)

<p>The slug technique proposed by Baird and Ho¹⁶ is a simple and easy way of obtaining the fundamental data on mass transfer coefficients in ternary liquid-liquid extraction. The technique yields a constant measurable interfacial area and the coefficients thus obtained from the experiments could be compared with those obtained by the penetration theory.</p> <p>The technique, as proposed, operates on a cylindrical slug of the organic phase suspended freely by downflow of water. This is modified in this work, by means of a disc arrangement to accommodate changes in aqueous phase flow rate. It was observed that the disc, which holds the slug at the center of the tube, has also enhanced the mass transfer rates by improving the mixing inside the slug, the improvement being higher for smaller slug lengths.</p> <p>The mass transfer coefficients in general were found to be lower than the penetration theory values. This could be due to any of the effects of dissolution of the solvent in the aqueous phase, hydrodynamic and surface conditions, variation of phase resistance with time, contaminants, and uncertainity about the distribution coefficient of the system: n-Heptane-Iodine-Water. Mass transfer rates in extraction with reactions were also found to be low and was thought to be due to violation of the assumption that the reaction between iodine and sodium thiosulphate is instantaneous. In the experiments with acetic acid as the solute, transfer rates were found to be higher than the expected because of Marangoni effects induced by the solute. It was concluded, however, that the technique will still be useful in obtaining the fundamental data with the use of a system with well-defined distribution data and redistilled solvents.</p> / Master of Engineering (ME)

<p>Polymer sintering can be described as the formation of a homogeneous melt from the coalescence of powder particles under the action of surface tension. Sintering is a fundamental phenomenon in processes such as rotational molding and powder coating. Rotational molding uses plastic powder to produce hollow plastic parts. The porosity of the final part produced in rotational molding depends on the completion of polymer sintering and the removal of bubbles. The objectives of this work are to study the effects of the material properties as well as the molding conditions on the sintering rate and to develop an appropriate model for polymer sintering.</p> <p>An experimental study of polymer sintering has been carried out. A reliable method has been developed for the obsevation and the measurement of the coalescence rate for two particles. The material viscosity, elasticity, and particle size were found to affect the sintering process. Most of the results obtained corroborate observations in rotational molding experiments. The effect of the particle geometry on the sintering rate was found to be negligible. This results has to led to the study of the rotomoldability of micropellets. Numerical studies have revealed that the processing conditions are severe and probably affect the micropellet rheology, which in return affects the coalescence process.</p> <p>A mathematical model describing the complete polymer sintering process has been developed. The approach was similar to that of Frenkel (1945). For Newtonian fluids, the proposed model's predictions are very close to Hopper's theoretical model (Hopper, 1984) and to numerical results for viscous sintering (Jagota and Dawson, 1988, Van de Vorst, 1994). The proposed model is successful in predicting the sintering rate for most of the rotational molding grade polyethylene resins used. However, all Newtonian models predict a faster coalescence rate than that observed with the copolymer resins used in this study. This result indicates that factors other than the surface tension and the viscosity play a role in polymer sintering.</p> <p>The proposed model has been generalized to describe sintering for viscoelastic fluids. As a first approach, the convected Maxwell constitutive equations were used together with the quasi-steady state approximation. The viscoelastic sintering model is capable of predicting the sintering rate observed in this study and the trends reported in the literature for the coalescence of acrylic resins.</p> <p>The combination of the present experimental and modeling studies can be used for the selection of appropriate materials and for improvement of the rotational molding process.</p> / Doctor of Philosophy (PhD)

<p>The value of model-based process optimization systems for competitive advantage in many industries, has been widely recognized. Such model-based optimization systems include Real-Time Optimization, On-Line Optimizing Control, off-line process scheduling, and any other economic process optimization scheme which uses a process model to predict optimal plant operation. The thesis investigates the design of these model-based optimization systems, particularly with respect to model structure and adjustable parameter selection.</p> <p>The main contribution of this work include design phase methods, based on fundamental principles of optimization and statistics theory, for determining whether a model-based optimization system can attain the plant optimum, as well as methods for discriminating between design alternatives. Three necessary conditions for zero-offset from the optimal plant operation are presented. These include Pont-Wise Model Adequacy, Augmented Model Adequacy and Point-Wise Stability. Recognizing that achieving zero-offset from the plant optimum may not always be possible, or may not be the only design objective, a Design Cost method is presented for selecting among design alternatives. This Design Cost method provides a natural "trade off" between offset elimination and variance of the predicted optimal manipulated variable values.</p> <p>Finally, the thesis is completed with a larger-scale case study involving the Williams-Otto Plant [1960]. In the case study selection of a process model and the adjustable parameter set for implementation in closed-loop Real-Time Optimization system is investigated.</p> / Doctor of Philosophy (PhD)

<p>The major complication limiting the usefulness of biomaterials in blood contracting situations is the activation of the coagulation pathways resulting in platelet activation and thromous formation at the surface. These events are initiated by the surface phenomena. The initial interaction is believed to be the adsorption of proteins onto the artificial surface followed by stimulation of cells and their subsequent adherence to the protein layer. The research presented in this thesis take the approach of making the surface typically interactive to achieve biocompatibility, and focuses on plasminogen, a plasma protein not previously studied in detail relative to blood interactions with artificial surfaces. The underlying working hypothesis of this research is that a potentially fibrinolytic, i.e. clot-lysing, surface may be formed via the specific and selective adsorption of plasminogen following blood contact.</p> <p>Our approach was to create materials that would selectively adsorb plasminogen from blood by incorporating lysine residues (known to bind plasminogen) into the surface. Several different materials have been developed to facilitate this investigation: polyurethanes substituted with lysine derived from sulphonated precursors, classical polyurethanes, and model silica glass materials fictionalized with sulphonic acid groups and subsequently derivatized with lysine.</p> <p>The model surfaces (silica glass, silylated sulphonated silica glass, and lysinated silica glass) were used to develop a method for the lysine derivatization of materials which results in incorporation of lysine at the surface with a free ε-amine group. Chemical and physical characterization of the surfaces utilizing ninhydrin reagent, contact angle and X-ray photoelectron spectroscopy were done and molecular modeling of silylated glass surface was undertaken. The data obtained indicate both qualitatively and quantitatively that the silylation and lysine fictionalization methods were successful.</p> <p>Protein adsorption from buffer and from citrated human plasma was investigated on all materials. Generally, the plasminogen adsorption from buffer was Langmuirian. Surface concentrations corresponding to monolayer protein adsorption were observed at the isotherm plateaus. The sulphonated and lysine derivatized materials showed higher adsorption capacities relative to other surfaces. Adsorption studies on the model materials from pure buffer, included both adsorption and the desorption experiments in the presence of competing species, and suggest that the binding mechanism of plasminogen is different on each of the surfaces. It seems likely that the lysine binding sites are involved at least to some extent in the adsorption of plasminogen to lysinated materials.</p> <p>The adsorption studies of fibrinogen and plasminogen from plasma suggest that plasminogen may be a possible contributor to the fibrinogen Vroman effect. In addition these studies provide further data in support of the contention that sulphonate groups have a strong influence on the adsorption of proteins in general. The plasma studies also show that the ratio of plasminogen to fibrinogen is greater on the lysinated than on the control on the sulphonated material, again indicating a specific influence of the lysine residues.</p> <p>Data on the enzymatic activity of plasminogen adsorbed on the lysinated silica glass and its sulphonated precursor are reported. These studies provide information on the effect of lysine derivatization on the very important question of plasmin activity of adsorbed plasminogen. It was found that the plasminogen is no "plasmin-like" when adsorbed to any of the surfaces studied in this system. However in the presence of plasminogen activator there is a highly significant difference in the enzymatic activity of plasminogen adsorbed to the lysinated materials and to its sulphonated precursor. The lysinated material shows two-four times the plasmin activity on a mole basis compared to the sulphonated material. The implications of these findings for the development of fibrinolytic surfaces are discussed.</p> <p>Preliminary investigations into polyurethanes derivatized using methods developed with the model materials are also reported.</p> / Doctor of Philosophy (PhD)