<p>In this study uninhibited styrene, degassed under high vacuum (10ˉ⁶ mm Hg), was thermally polymerized in bulk of 10ˉ⁶ mm Hg to almost complete conversion. The samples so obtained were thermally degraded under the same high vacuum conditions at temperatures ranging between 270°C and 340°C for different time intervals over the whole range of degradation. The breakdown of the polymer, through changes in number, weight average molecular weights, polydispersity and molecular weight distribution, was investigated using gel permeation chromatography. A mechanism of degradation was sought, and the breakdown explained by a combined process of deploymerization and chain scission. Scission constants were determined and compared with the constants, obtained from a polystyrene having much larger molecular weights.</p> / Master of Engineering (ME)

<p>Uranium-bearing ore is crushed and leached to extract uranium in an ore processing plant. However, the decay products of uranium are discarded as wastes or tailings from the plant. One such decay product is Ra²²⁶, which is quite toxic and mobile. It remains dispersed with the tailings and spreads into air and potable waters from the tailing areas due to natural forces and decay to gaseous Rn²²².</p> <p>In this work, bench scale leaching of the uranium mill tailings was done with a view to extract Ra²²⁶ into the liquid phase. A gamma counting technique for determination of Ra²²⁶ concentrations in solid and liquid samples was also developed. Details of sample preparation, ingrowth, and calibration are discussed. For extraction of radium, several extractants such as water and solutions of CaCl₂, EDTA, DTPA and HNO₃ were used. Effects of solid to liquid ratio and time of contact on the extraction of Ra²²⁶ were studied for these extractants. In some cases, effects of temperature, concentration of extractant and multiple cross current contacting were also studied. The results of various leaching experiments are interpreted.</p> <p>The main results show that equilibrium distribution of radium between the extractants and tailings was attained rapidly. In all experiments, a fraction of radium was "unextractable", i.e., all operating lines had an intercept when extrapolated to zero solid to liquid ratio. In general, the operating lines had a relatively small slope indicating that high solid to liquid ratio could be used without significantly degrading extraction.</p> / Master of Engineering (ME)

<p>A key unit operation in most recycled paper mills is the flotation deinking process. In an ideal flotation cell, the froth (rejects) should only contain ink and other hydrophobic contaminants. In actual flotation cells, however, the rejects also contain small amounts of useful cellulosic pulp material. These pulp losses are a waste of raw material. The objectives of this thesis were to identify the dominant mechanism(s) responsible for pulp loss in flotation systems, and to attempt to quantify these mechanisms. Three potential causes of pulp loss were investigated: pulp flotation by bubble adhesion, pulp flotation by bubbles mechanically-trapped within fibre networks, and the transport of pulp fibres and pulp fines into the froth by hydraulic entrainment. Direct observation of bubble-fibre interactions in both bubble generation tests and in flow visualization studies revealed that bubbles are unlikely to adhere to wetted pulp fibres. The flow visualization studies also demonstrated that while bubbles can become trapped in fibre networks, these bubbles are easily released under flowing conditions. Mechanical entrapment is considered to cause significant pulp flotation only under quiescent conditions where pulp flocs were present. Based on results obtained from batch flotation experiments, it was determined that hydraulic entrainment is the dominant pulp loss mechanism. Entrainment was largely affected by flocculation, as rising bubbles tended to be channeled around the flocs. Two regimes of entrainment were proposed, which depended on flocculation. When no flocs were present, the consistency of the foam and feed were similar. In flocculated pulp suspensions, the consistency of the foam represented that of the pulp existing in regions between the flocs. A three-parameter model was derived which predicted the consistencies and the pulp fines fractions of the foams from the flotation experiments. By combining these experimental results with independently-measured pulp sedimentation curves, the model characterized the pulp suspensions' state of flocculation in terms of floc volume fraction, floc consistency and the distribution of pulp fines between the flocculated and non-flocculated regions. The model was more applicable when flow conditions were such that small bubbles were channeled around the flocs. The model also indicated that pulp fines tend to be excluded from flocs.</p> / Doctor of Philosophy (PhD)

<p>The displacement of aqueous Kraft model black liquor (MBL) from a porous bed with cationic polymer solution (polyDADMAC) was investigated to determine the mechanisms by which polyDADMAC influenced the miscible displacement. The porous bed consisted of a 1.7 cm diameter central channel of 638 μm glass beads surrounded by an annulus of 121 μm glass beads contained in a 5.2 cm internal diameter cylindrical glass column. Flow visualization experiments and displacement washing experiments were performed. Conductivity probes were installed at twelve locations inside the beds to monitor the velocity and locations of displacement fronts. Flow visualization experiments in a transparent cell gave direct evidence of the presence of cross flow of fluid elements from annulus to the channel in a model channel bed during water displacement. After selective plugging of the channel with precipitate formed from reaction of lignin and polyDADMAC, fluid elements in annulus were observed to move straight down the model channel bed. Results of the displacement experiments in the model beds confirmed that washing efficiency, a measure of the miscible displacement performance, increased by 1.7 to 2 times and permeability of channel decreased by 34% to 18% when the bed was displaced with polymer solution instead of water. The improvement in the displacement performance with polymer solution resulted due to the reduction in channeling of polymer solution in the model channel bed. Results of data obtained by the probes gave direct evidence of the reduction in velocity and mixing length of front in the channel confirming reduction in the channeling in a model channel bed during displacement with polymer solution. Increased velocity of front in the annulus during displacement with polymer solution confirmed better miscible displacement in a channel bed with polymer solution than that with water. A communicating channel bed model was developed based on a technique called network of zones. The model predicts the breakthrough curves and the profiles of interstitial velocities inside the model beds during different displacement washing conditions. The model successfully predicts pressure drop profiles across homogeneous beds during displacement with either water or polymer solution and channel bed during washing with water. However, the pressure drop predicted by the model was higher than that obtained experimentally during washing with polymer solution in a model channel bed.</p> / Doctor of Philosophy (PhD)

<p>This work examines the relationship between dissolved oxygen level and the nitrification process in municipal wastewaters. Carbon removal, chemical oxygen demand removal and sludge viability were also investigated. Parallel bench scale continuous activated sludge reactors operating at dissolved oxygen levels of approximately 2 mg/l and 8 mg/l were used.</p> <p>The high dissolved oxygen level sludge produced significantly more nitrate nitrogen and exhibited significantly greater nitrification rates overall. The nitrification rate differential increased with temperature. The nitrification rate appeared to be more sensitive to temperature and sludge age at the higher dissolved oxygen level.</p> <p>Significantly greater overall TOC reduction and TOC removal rates were observed at the high dissolved oxygen-level although no one set of experimental conditions exhibited a significant difference at the 99% confidence level. No significant difference in effluent COD quality or removal rate was observed.</p> <p>There was a significant difference at the 95% confidence level in unit ATP levels, indicating that the high dissolved oxygen level sludge may have been more viable.</p> / Master of Engineering (ME)

<p>Fault Detection and Diagnosis have become important topics in the process industries. The off-line diagnosis of past transient upsets can lead to important process or operation modifications that can improve the future behavior of the process. The rapid on-line diagnosis of faults is even more important since it can anticipate and minimize the impact of otherwise costly effects.</p> <p>The first part of this thesis addressed the problem of fault diagnosis in multivariate, dynamic, continuous chemical processes. Two types of faults were considered: deterministic (whose root cause is a randomly occurring deterministic event) and stochastic (caused by an underlying stochastic process). A realistic simulation of a chemical plant was used as a test bed for the proposed methods. Due to the lack of accurate dynamic models for this type of process, a Pattern Recognition approach was followed. Within this framework, several methods were designed for the on-line and offline diagnosis of both types of faults. All methods consisted of: I) a feature extraction step, where magnitude invariant features are extracted from both the reference patterns and the pattern of the new unknown fault, and II) a similarity assessment step where the distance between the new pattern and each of the reference patterns is estimated using Dynamic Time Warping.</p> <p>Due to the use of magnitude invariant features and the ability of Dynamic Time Warping to synchronize similar patterns with distorted temporal correlations, the results were satisfactory in diagnosing deterministic faults. In the case of stochastic faults, the results were inconclusive. The correlation pattern between the variables was used as the feature for the diagnosis of stochastic faults. However, the slow dynamics and the effect of the recycle in the simulated chemical plant meant that unrealistically long records of data are required for an accurate estimate of this feature.</p> <p>The second part of the thesis investigated the problem of fault detection in batch processes, and in particular the problem of batch trajectories of unequal duration and poor synchronization. A new method, based on Dynamic Time Warping, was proposed for the synchronization of batch trajectories of this type; the method is multivariate and requires minimal process knowledge. It was also shown how to use Dynamic Time Warping to synchronize a new batch trajectory with the reference trajectories so that batch monitoring methods based on Multivariate Statistical Methods could be used. Finally, a new on-line monitoring method was presented, based on the concept of instantaneous quadratic distance, which does not require prediction of the future behavior of the batch trajectory.</p> / Doctor of Philosophy (PhD)

<p>Reaction injection moulding (RIM) is commercially used in industry for making polymer parts. In this process, two or more jets produce impingement mixing of low viscosity monomers or oligomers in a mixhead. In order to better understand an impingement mixer, it is essential to understand the flow field created by the opposed laminar jets in a similar configuration. In this thesis, an experimental investigation and analysis of the flow field created by the two equal and opposed cylindrical laminar jets impinging near the closed end of a confined space (mixhead) has been conducted. Three important flow and geometrical parameters which influence the flow field are considered. These are (i) jet Reynolds numbers (Re), (ii) H/D, a ratio of the position (H) (the distance in - Z direction from the impingement point to the closed end) to mixhead diameter (D), and (iii) viscosity of the fluid used. Both qualitative and quantitative studies have been performed using an experimental model of the mixhead which simulates a RIM type configuration. For qualitative analysis, two flow visualization methods, namely, dye injection and particle tracing methods have been used. For quantitative analysis, the laser Doppler anemometer (LDA) technique has been used to quantify the velocity fields (only U and V components).</p> <p>It has been found that Re shows a significant effect on the flow field. For Re< 50, flow remains in a statically stable mode and rotating vortices are formed near the impingement plane. At slightly higher Re (≥50), the flow pattern changes to dynamically stable mode and the impingement plane oscillates. At Re≥90, the flow attains an unstable mode. In fact, the vortices gradually grow with increasing Re and at a critical value of Re they dynamically interact with each other causing oscillations about the impingement plane. For both static and dynamic stable zones, the length of circulating region increases as a function of Re (l/D = a - be⁻ᶜᴿᵉ) between 50 to 90. For Re > 90, the recirculating region shrinks before breaking into an unstable zone at a Re of approximately 125.</p> <p>The piston position (H/D) plays an important role in the stability of the flow pattern. An unstable flow pattern transforms to a steady flow pattern when the piston is moved down from H/D = -0.5 to -0.75 at Re = 150. Furthermore, it appears that the flow patterns are always unstable at very high Re values since no stable patterns could be obtained up to HID = -1.0 at Re = 200. In stable modes, flow fields of both side of impingement plane are like a mirror image of each other. It appears that viscosity of the fluid has no significant influence on the flow patterns at constant Re. The vector and streamline plots obtained by LDA for Re = 50, and 150 agree with the flow pattern obtained with flow visualization methods. Also, the experimental measurements confirm that the flow in the vortex regions is 3-dimensional.</p> <p>Vigorous fluctuations are evident up to one-half mixhead diameter above the impingement point. The axial velocity (V) along the mixhead axis reaches a velocity profile of a developed one dimensional steady flow within 1 to 1.5 mixhead diameters. However, the radial velocity (U) along the mixhead axis reduces to an insignificant value at a distance equal to approximately two jet-diameters above the impingement point. The frequency analysis of the time series of velocity components showed higher harmonics as the Re increases from Re = 50 to 150. The periodic structure completely vanishes at Re = 200. The periodic nature of time series which vanishes at Re = 200 for H/D = 0, vanishes at Re = 150 when H/D = -0.5. This shows that H/D ratio also influences the breakdown of any periodic nature in the flow field.</p> / Doctor of Philosophy (PhD)

<p>Membrane surface modification is of interest to most membrane scientists because the surface properties of most commercial membranes are often nonoptimal for particular applications. This thesis focuses on the further development of a thin film coating on the internal surface of polypropylene microfiltration membrane. A modified interfacial polymerizatIon of 1,6-naphthalene disulfonylchloride containing a photoactive diazoketone (DK) group on the side chain and 1,8-octanediamine is studied. Various fabricatIon conditions including concentration of monomers, organic solvent selection, crosslinking agent, selection of an acid acceptor and reaction time were investigated. The choice of organic solvent was found to be critical for obtaining an even, smooth coating layer without significantly changing the membrane morphology. Mass of the polysulfonamide coating layer was typically 10-15 wt% of the base membrane. The coating layer at the membrane internal surface demonstrated a high stability to Soxhlet extracting using CHCl₃ for 48 hours.</p> <p>A mechanism of this new type of interfacial polymerization is qualitatively proposed, and investigated experimentally. The first reported study of end-group analysis and molecular weight measurement of coating polysulfonamide provided a solid ground to evaluate the hypothesis. Most results from experiments either strongly support the proposed mechanism or can be reasonably interpreted by the proposed mechanism. The mechanistic study defines the most important principles of this new type of interfacial polymerization which can be varied to control the coating process, and gives a better understanding of this new coating technology.</p> <p>General applicability of this coating technology was tested by applying this technology to polyester and polyamide. The results indicate that this new technology is generally applicable and can be used to coat condensation polymers to the internal membrane surface as an even thin layer, without significantly changing the morphology of the polyolefin microfiltration membranes.</p> <p>The chemical functionality of polysulfonamide coated membrane surface could be further modified via photochemical transformation of the diazoketone group contained in the coating polysulfonamide, in the presence of an appropriate medium. Thus, membranes containing indene acid, glycolic acid, bromoethyl ester, and polyethylene glycol ester functionalities were obtained. The photochemical reaction conversion was found to be approximately 80%. The highest charge density (0.097 meq/g) could be achieved on the glycolic acid ester membrane. The hydrophilicity of the membrane surface is increased by polysulfonamide coating and further increased by photochemical functionalization.</p> <p>The microfiltration flux and separation performance of these membranes with polystyrene latex (PSL) and carboxylate modified latex (CML) spheres, was examined. The extent of fouling of the membranes was studied by SEM. A lower separation of PSL and a higher pure buffer flux were found for the functionalized membrane compared to the polysulfonamide coated control (DK) membrane. Pure buffer flux was increased after photochemical modification. These results suggest that the increase of hydrophilicity of the photochemically modified membranes was the dominant factor affecting the membrane separation performance. The degree of fouling of the functionalized membranes was significantly decreased which would presumably lead to an improvement of membrane effective life-time.</p> <p>A remarkable change of polypropylene microfiltration membrane performance was found after the addition of Triton X-100 to the PSL feed solution. This phenomenon was explained in terms of the change of membrane surface properties. Polyethylene glycol ester membranes were introduced to mimic the membrane physically adsorbed Triton X- 100. The results suggested that the Triton X-100 in the testing solution not only temporarily modified the membrane surface properties to be more hydrophillic, but also prevented the PSL particles from aggregating, particularly in the membrane pores.</p> / Doctor of Philosophy (PhD)

<p>Therein is reported an experimental evaluation of a model after Abdel-Alim and Hamielec, for bulk and suspension polymerization of vinyl chloride for the conditions of nonisothermal polymerization and polymerization with multiple initiator systems.</p> <p>The model predictions for these rather extreme conditions are reasonable. It appears that with minor modifications the model could be used to design initiator systems and to develop new molecular weight recipes for commercial initiators.</p> / Master of Engineering (ME)

<p>The limiting current method was used to investigate the mass transfer to a wall of a rectangular Plexiglass channel operated as a flow reactor for Reynolds numbers ranging from 200 to 32000. Cathode and mode blocks with smooth nickel surfaces were mounted in opposite walls of the channel, flush with the interior surface. Isothermal and non-isothermal mass transfer experiments were performed with the entire horizontal flow channel</p> <p>i. rotated in 30° increments so that the cathode position varied from facing up to facing down (angle α) ii. tilled in 30° increments so that the channel position varied from horizontal to vertical (angle β)</p> <p>at various bulk solution and cathode surface temperatures.</p> <p>lsothermaI experiments (3277 in total) were performed in a flow channel with an internal rectangular cross-section of 2 cm x 1.5 cm (width x height), the total length being 229 cm between inlet and outlet. The electrolyte bulk solution, comprised of 2 mol/l NaOH, 0.025 mol/l K₃Fe(CN)₆ and 0.025 mol/l K₄Fe(CN)₆, was varied in temperature from 25°C to 55°C.</p> <p>While the bulk solution was maintained at 25°C and the cathode surface temperature varied from 35°C to 55°C, non-isothermal experiments (2394 in total) were conducted in the same apparatus as describe above.</p> <p>Examination of the data revealed that the mass transfer in the laminar now regime, at Re < 1900, seems to be independent of the Reynolds number. For laminar now conditions an equation of the form</p> <p>Sh = 0.116 (1 + 0.336 (sinα)²⁷⁶⁶ + 0.427 (sinβ)⁰⁸²⁶(GRmSc)⁰³²⁰</p> <p>gives an adequate prediction for mass transfer rates at 0° ≤ α ≤ 180° and 0° ≤ β ≤ 60°, where GRm is the combined Grashof number for heat and mass transfer. Mass transfer rates for the channel in a vertical position (β = 90°) can be predicted by:</p> <p>[equation removed]</p> <p>For Reynolds numbers greater than 1900, mass transfer was found to be controlled by free and forced convection, with the forced convection dominant. For the turbulent flow region the date for 0° ≤ α ≤ 180° and 0° ≤ β ≤ 60° can be predicted by:</p> <p>[equation removed]</p> <p>Mass transfer rates for channel slopes of 90° were best correlated by:</p> <p>[equation removed]</p> <p>The isothermal laminar and turbulent experimental mass transfer coefficients were found to be within ±10% of correlations presented by Rousar et al. (1971) and Pickett (1977), respectively. For non-isothermal mass transfer in a horizontal channel under laminar and turbulent flow conditions the correlations of Ajersch (1990) agree well with the experimental data. Non-isothermal operation was found to increase mass transfer approximately four times over the isothermal case under the optimum enhancement conditions.</p> <p>Since the system under investigation is a couple one, no analytical solution of the growing equations is possible. The finite element solver FIDAP was used to obtain numerically the local mass transfer coefficients, local Reynolds numbers, local densities, stream functions and concentration and temperature profiles for isothermal and non-isothermal case studies. The numerically obtained average mass transfer coefficients agreed within 15% of the experiment values for both the isothermal and non-isothermal case.</p> / Doctor of Philosophy (PhD)