Process development of new complex reactions in the pharmaceutical and fine chemicals industries is challenging, and expensive. The field is beginning to see a bridging between fundamental first-principles investigations, and utilisation of data-driven statistical methods, such as machine learning. Nonetheless, process development and optimisation in these industries is mostly driven by trial-and-error, and experience. Approaches that move beyond these are limited to the well-developed optimisation of continuous variables, and often do not yield physical insights. This thesis describes several new methods developed to address research questions related to this challenge. First, we investigated whether utilising physical knowledge could aid statistics-guided self-optimisation of a C-H activation reaction, in which the optimisation variables were continuous. We then considered algorithmic treatment of the more challenging discrete variables, focussing on solvents. We parametrised a library of 459 solvents with physically meaningful molecular descriptors. Our case study was a homogeneous Rh-catalysed asymmetric hydrogenation to produce a chiral γ-lactam, with conversion and diastereoselectivity as objectives. We adapted a state-of-the-art multi-objective machine learning algorithm, based on Gaussian processes, to utilise the descriptors as inputs, and to create a surrogate model for each objective. The aim of the algorithm was to determine a set of Pareto solutions with a minimum experimental budget, whilst simultaneously addressing model uncertainty. We found that descriptors are a valuable tool for Design of Experiments, and can produce predictive and interpretable surrogate models. Subsequently, a physical investigation of this reaction led to the discovery of an efficient catalyst-ligand system, which we studied by operando NMR, and identified a parametrised kinetic model. Turning the focus then to ligands for asymmetric hydrogenation, we calculated versatile empirical descriptors based on the similarity of atomic environments, for 102 chiral ligands, to predict diastereoselectivity. Whilst the model fit was good, it failed to accurately predict the performance of an unseen ligand family, due to analogue bias. Physical knowledge has then guided the selection of symmetrised physico-chemical descriptors. This produced more accurate predictive models for diastereoselectivity, including for an unseen ligand family. The contribution of this thesis is a development of novel and effective workflows and methodologies for process development. These open the door for process chemists to save time and resources, freeing them up from routine work, to focus instead on creatively designing new chemistry for future real-world applications.

Parkinson’s Disease (PD) is a progressively debilitating neurodegenerative disease that affects the central nervous system and leads to severe difficulties with body movements. PD occurs due to the selective degeneration of neurons in the region of the brain known as the substantia nigra pars compacta. To date PD remains an incurable disease. Currently prescribed drugs provide only symptomatic relief to patients. Furthermore, they have considerable side effects and are often ineffective in the later stages of the disease or need to be used in combination in order to be effective. The role of neuroprotectants as a preventative measure in PD therapy is consequently receiving a great deal of attention and is being subjected to extensive research. This study sought to develop a novel prolonged-release drug delivery device for providing site-specific administration of newly researched neuroprotective agents. Nicotine was employed as the model neuroprotectant to incorporate in a novel reinforced crosslinked multiple-unit multi-polymeric drug delivery device. The study was the first of its kind to develop and employ the alkaloid for this purpose in a formulated delivery system. The device was intended to be one that provided zero-order prolonged release of the drug over a period of 1-2 months. The device was formulated such that its design was in keeping with the potential for implantation into the substantia nigra pars compacta to provide site-specific drug delivery. In order to do so, polymers, with biocompatible and bioerodable characteristics were selected to incorporate the drug within a reinforced crosslinked matrix. The study elucidated the mechanism of crosslinking of ionotropically crosslinked alginate spheres (gelispheres) with a variety of crosslinking agents through an evaluation of physicomechanical properties of the crosslinked system. The presence of barium in the crosslinked matrices generated densely networked gelispheres which retained their robustness following exposure to hydrating media and displayed promising potential for 4 entrapping drug molecules and retarding their release. The system was reinforced employing hydroxyethylcellulose (HEC) and polyacrylic acid (PAA). A Design of Experiments approach employing a Plackett-Burman screening design enabled optimisation of the proposed device in terms of reinforcing polymers (HEC and PAA) and crosslinking agents (barium and calcium). In order to further attenuate drug release rate the optimised crosslinked gelispheres were exposed to dilute hydrochloric acid (HCl) which significantly decreased gelisphere matrix swelling and erosion following exposure to simulated cerebrospinal fluid (CSF). These gelispheres were thereafter incorporated into a compressed release-rate modulating polymeric discs. Zero-order drug release was observed for a period of 50 days in simulated CSF when the optimised gelispheres were incorporated into compressed poly(lactic-co-glycolic) acid (PLGA) discs. Alternative approaches to modify drug release kinetics were also evaluated including the use of PLGA coatings on compressed hydroxypropylmethylcellulose-polyethylene oxide (HPMC-PEO) discs incorporating gelispheres and the use of crosslinked alginate-pectinate gelispheres as carrier systems to deliver PLGA-PLA (polylactic acid) microparticles incorporating drug. A Box-Behnken statistical design was employed to formulate and optimise the drug carrying PLGA-PLA microparticles. In both of the abovementioned cases we obtained sustained zeroorder drug release.

<p>Glycerol is one of the by-products of transesterification of fatty acids for the production of bio-diesel. Value-added products such as hydrogen, wood stabilizers and liquid chemicals from catalytic treatment of glycerol can improve the economics of the bio-diesel production process. Catalytic conversion of glycerol can be used for production of value-added liquid chemicals. In this work, a systematic study has been conducted to evaluate the effects of operating conditions on glycerol conversion to liquid chemical products in the presence of acid catalysts. </p><p>Central composite design for response surface method was used to design the experimental plan. Experiments were performed in a fixed-bed reactor using HZSM-5, HY, silica-alumina and ã-alumina catalysts. The temperature, carrier gas flow rate and weight hourly space velocity (WHSV) were maintained in the range of 350-500 oC, 20-50 mL/min and 5.40-21.60 h -1, respectively. </p><p>The main liquid chemicals detected in liquid product were acetaldehyde, acrolein, formaldehyde and hydroxyacetone. Under all experimental conditions complete glycerol conversion was obtained over silica-alumina and ã-alumina. A maximum liquid product yield of approximately 83 g/100g feed was obtained with these two catalysts when the operating conditions were maintained at 380 oC, 26 mL/min and 8.68 h-1. Maximum glycerol conversions of 100 wt% and 78.8 wt% were obtained in the presence of HY and HZSM-5 at temperature, carrier gas flow rate and WHSV of 470 oC, 26 mL/min and 8.68 h-1. HY and HZSM-5 produced maximum liquid product of 80.9 and 59.0 g/100 g feed at temperature of 425 and 470 oC, respectively.</p><p>Silica-alumina produced the maximum acetaldehyde (~24.5 g/100 g feed) whereas ã-alumina produced the maximum acrolein (~25 g/100 g feed). Also, silica-alumina produced highest formaldehyde yield of 9g/100 g feed whereas HY produced highest acetol yield of 14.7 g/100 g feed. The effect of pore size of these catalysts was studied on optimum glycerol conversion and liquid product yield. Optimum conversion increased from 80 to 100 wt% and optimum liquid product increased from 59 to 83.3 g/100 g feed when the pore size of catalyst was increased from 0.54 in case of HZSM-5 to 0.74 nm in case of HY, after which the effect of pore size was minimal.

While designing jet-engine components information about the loads that the component will be subjected to, is critical. For this, a full system analysis of the engine is often performed with every component put together in a large finite element model, which is called the whole engine model (WEM). This model will mostly be composed of lower order shell elements with a few thousands of elements. At the design level of a component, the FE model is detailed with several hundred thousand higher order solid elements. The detailed model cannot be directly put into the whole engine model due to excessive run times. Therefore there must be a simpler representation of the component -a sub-model- with much fewer elements so that it can be assembled into the whole engine model. This simple model must have the same stiffness (load/displacement) in chosen directions, the same mass and the basic mode shapes and frequencies should also be the same with the detailed model. 4 different structural optimisation schemes were studied to prepare a model: sizing optimisation, optimisation with material properties as design variables, combined sizing and material property optimisation and free-size optimisation. Among these free-size optimisation where each element in the model has its own design variable -thickness- was found to be the most effective method. The stiffness could be matched to the detailed model as close as 5% and so also could the first two fundamental mode shapes and frequencies. Additionally, the initial sub-model prepared was used to do a preliminary study on how variations in casting dimensions would affect the stiffness of the component in a certain direction. This was done by creating a design of experiments (DoE) for the stiffness. A response surface for the stiffness was created in terms of the dimensions that have the most significant effect. This was later used to predict an interval for the stiffness based on variations in the cast dimensions with a confidence level of 99.7%.

<p>Glycerol is one of the by-products of transesterification of fatty acids for the production of bio-diesel. Value-added products such as hydrogen, wood stabilizers and liquid chemicals from catalytic treatment of glycerol can improve the economics of the bio-diesel production process. Catalytic conversion of glycerol can be used for production of value-added liquid chemicals. In this work, a systematic study has been conducted to evaluate the effects of operating conditions on glycerol conversion to liquid chemical products in the presence of acid catalysts. </p><p>Central composite design for response surface method was used to design the experimental plan. Experiments were performed in a fixed-bed reactor using HZSM-5, HY, silica-alumina and ã-alumina catalysts. The temperature, carrier gas flow rate and weight hourly space velocity (WHSV) were maintained in the range of 350-500 oC, 20-50 mL/min and 5.40-21.60 h -1, respectively. </p><p>The main liquid chemicals detected in liquid product were acetaldehyde, acrolein, formaldehyde and hydroxyacetone. Under all experimental conditions complete glycerol conversion was obtained over silica-alumina and ã-alumina. A maximum liquid product yield of approximately 83 g/100g feed was obtained with these two catalysts when the operating conditions were maintained at 380 oC, 26 mL/min and 8.68 h-1. Maximum glycerol conversions of 100 wt% and 78.8 wt% were obtained in the presence of HY and HZSM-5 at temperature, carrier gas flow rate and WHSV of 470 oC, 26 mL/min and 8.68 h-1. HY and HZSM-5 produced maximum liquid product of 80.9 and 59.0 g/100 g feed at temperature of 425 and 470 oC, respectively.</p><p>Silica-alumina produced the maximum acetaldehyde (~24.5 g/100 g feed) whereas ã-alumina produced the maximum acrolein (~25 g/100 g feed). Also, silica-alumina produced highest formaldehyde yield of 9g/100 g feed whereas HY produced highest acetol yield of 14.7 g/100 g feed. The effect of pore size of these catalysts was studied on optimum glycerol conversion and liquid product yield. Optimum conversion increased from 80 to 100 wt% and optimum liquid product increased from 59 to 83.3 g/100 g feed when the pore size of catalyst was increased from 0.54 in case of HZSM-5 to 0.74 nm in case of HY, after which the effect of pore size was minimal.

This research details the development of a novel process for four commercially available no-flow fluxing underfills for use with flip chip on FR4 substrates. The daisy chain test die was used such that two point resistance measurements could be used to determine the integrity of the solder interconnects post reflow. The impact of the underfill dispensing pattern on underfill void formation is determined in a full factorial dispense DOE that includes two factors: pattern and speed. Evaluation metrics include underfill material voiding and fillet shape. The impact of the placement process is determined in a second full factorial DOE involving three factors at two levels each: dispense pattern, placement force, and dwell time. Metrics include interconnect yield and underfill voiding. The results of these DOEs are used to select an optimal placement process for each material to be used for the remaining reflow experiments. The process developed is a novel approach to no-flow processing; the material is dispensed to the side of the bond site and allowed to flow under the chip after placement by capillary action during the early stages of reflow. This development allows for void free assemblies using no-flow materials. Reflow parameters are investigated using a parametric approach. The following parameters are varied at 2 levels individually off a baseline profile: Peak Temperature, Time > 183 oC, Peak Ramp Rate, Soak Time, and Soak Temperature. A ranking was developed for each material based on the observable metrics: interconnect yield, underfill material voiding, two point resistance, and a grain area fraction term. The results were used to select an optimal assembly process for each material. Test boards were assembled in replicates of 30 according to the optimal process for each material, and AATC -40 to 125 oC thermal cycling test was performed. The MTTF for these assemblies has exceeded 3000 cycles; the void free process successfully avoids premature failure due to solder extrusion into voids. Further process development work has demonstrated that the process is scalable to larger area array die and other edge dispense patterns have also been demonstrated to result in void free assemblies.

Fossil fuels and their byproducts are a vital part of our economy, and society. Until renewable energy sources and energy storage technologies advance to the point where they are reliable and inexpensive, the US Economy will continue to depend upon fossil fuels. Current resources are being consumed, and the "easy to reach" reserves are becoming depleted. This leads to the requirement for more exploratory drilling, and the potential for more disasters like the recent Deepwater Horizon spill in the Gulf of Mexico. Drilling is the first of several steps in the creation of a productive oil or natural gas well. Completing a well involves casing the walls in concrete to prevent damage to the surrounding rock formations and to ensure that all of the oil or gas is captured without escaping to the surrounding environment. Ensuring the piping, which is used to case wells, is assembled correctly and to manufacturer's specifications is the focus of this study. Individual pipe sections are screwed together with a requirement for torque and number of turns. Each joint must be verified to ensure integrity, and minimize the possibility of a spill or leak. The torque measurement can be accomplished by a "torque sub", a sensor installed in-line with the drill string. The torque sub is a wireless sensor that transmits torque data to the control system for logging and display. This thesis defines the parameters required to integrate a "number of turns" measurement into an existing torque sub so that both parameters can be captured, recorded and reported using a single device. The Yost Engineering 3-Space Sensor was evaluated for use in this application. The configuration that gave the most accurate data was selected, along with the determination of some correction factors to account for site specific variation in the signals. A calibration algorithm is discussed, along with several unique methods for ensuring that the sensor output doesn't drift over the course of the joint make-up process. / text

No description available.

Emergency department(ED) is one of the busiest clinical units in Winnipeg Victoria Gen-eral Hospital (VGH) which faces the challenge of patients’ long waiting-time as increas-ing healthcare demand and limited resources. This research investigates the critical factors of the ED operation to enhance the operational efficiency using simulation modeling and optimization. The contribution of this research is the integration of simulation and optimization for the performance improvement of ED operations. Discrete-events simula-tion (DES) methodology provides a cost-effective tool to analyse the performance of the ED operations and evaluates the potential alternatives. Design of experiments (DOE) and Scatter search (SS) of model optimization are proposed to search the ED potential capaci-ty for waiting-time reduction. The patient-flow is accelerated along with the waiting-time reduction, which results in better efficient patient throughput in the ED. A specific strate-gy is suggested to improve the ED operation based on the simulation model.

We investigate various methods for testing whether two groups of curves are statistically significantly different, with the motivation to apply the techniques to the analysis of data arising from designed experiments. We propose a set of tests based on pairwise differences between individual curves. Our objective is to compare the power and robustness of a variety of tests, including a collection of permutation tests, a test based on the functional principal components scores, the adaptive Neyman test and the functional F test. We illustrate the application of these tests in the context of a designed 2^4 factorial experiment with a case study using data provided by NASA. We apply the methods for comparing curves to this factorial data by dividing the data into two groups by each effect (A, B, . . . , ABCD) in turn. We carry out a large simulation study investigating the power of the tests in detecting contamination, location, and shift effects on unimodal and monotone curves. We conclude that the permutation test using the mean of the pairwise differences in L1 norm has the best overall power performance and is a robust test statistic applicable in a wide variety of situations. The advantage of using a permutation test is that it is an exact, distribution-free test that performs well overall when applied to functional data. This test may be extended to more than two groups by constructing test statistics based on averages of pairwise differences between curves from the different groups and, as such, is an important building-block for larger experiments and more complex designs.