This thesis is a study of the chemical and physical changes which can occur to gas turbine lubricants as a result of exposure to operational conditions. The continual evolution toward more efficient gas turbines is accompanied by increasing thermal and mechanical loading which the lubricant must be able to withstand. In this thesis two major degradation issues are studied; thermal oxidative degradation and lubricant deposition. In the area of thermal oxidative degradation, efforts are made to better understand the key parameters which determine the lubricant breakdown mechanism. Through control of these parameters and comparison to service derived gas turbine oil samples a new laboratory methodology is proposed for the assessment of lubricant oxidative degradation. The study of lubricant deposition in this thesis is concentrated on the regions of highest risk, the bearing chamber feed (single phase) and vent (two phase) oil pipes. Development of existing laboratory scale deposition simulators was conducted to increase how engine representative the methods are of gas turbine conditions. These simulators were used to evaluate the rate of deposition with a range of lubricants, simulated engine cycles and pipe surfaces.

The spray and combustion characteristics of the new bio-fuel candidates, 2,5-dimethylfuran (known as DMF) and 2-methylfuran (known as MF), are examined using optical diagnostic methods. A macroscopic spray characteristics study using high speed imaging has been performed to gain the understanding of the bio-fuels’ spray behavior under various conditions compared to gasoline and isooctane. The droplet sizes of the bio-fuel injections under different operating conditions are also studied using Phase Doppler Particle Analyzer (PDPA). The laminar flame propagation and the turbulent flame propagation for MF and DMF have been benchmarked against isooctane.

From the plethora of possible microbial hosts, Escherichia coli remains the most widely used “microbial factory” for recombinant protein production (RPP). However, despite the numerous advances, RPP in E. coli is still a significant challenge. Strong promoters to achieve high expression and protein production levels are often used, however, the underlying effects on the host physiology are often unseen. This thesis reports the development of “stress-minimisation” approaches for the production of recombinant proteins, either targeting the cytoplasm or the periplasmic space. First, the fermentation conditions for the production of TNFα in the cytoplasm were optimised. The expression vector, culture medium, temperature, inducer concentration and induction point were optimised, yielding 5.35 g ∙ L-1 of rhTNFα, 70% being accumulated as a soluble product. Second, different approaches for the production of recombinant proteins targeting the periplasm were evaluated, using an antibody fragment, scFv163R4, as a model protein. The effect of different growth conditions and signal peptides on the production of the scFv163R4 were evaluated. However, the selection of the optimal signal peptide was proven to be challenging, establishing the requirement of a high-throughput screening assay. This resulted in the development of a screening assay using β-lactamase as a reporter protein, for the evaluation of mutant signal peptide libraries with improved translocation activity. Initial evaluations resulted in the selection of two mutant signal peptides with enhanced translocation of scFv163R4-A, yielding almost 1 g ∙ L-1 of periplasmic scFv163R4-A. The generation of mutant signal peptide libraries in combination with the β-lactamase screening assay represents an important advance for the production of disulphide bonded proteins for the biotechnology industry.

This PhD project aimed to study the separation of carbon dioxide from high pressure gas mixtures as it is directly applicable to pre-combustion carbon dioxide capture. Adsorption isotherms were produced and best fit by the Langmuir-Freundlich and dual-site Langmuir (DSL) isotherms. Breakthrough experiments investigated the separation under dynamic conditions. These showed that adsorption capacities need to be studied on a volumetric basis. The multicomponent DSL isotherm model was the most suitable for predicting breakthrough capacities. An axial dispersed plug flow model was validated against the experimental data with a reasonable accuracy. Cyclic experiments were validated, which were found to be restricted by the surrounding pipework and instruments. A parameter sensitivity analysis indicated the particle diameter, bed voidage and particle voidage had the greatest effect on the breakthrough curve. Pressure swing adsorption systems were also simulated. Simple cycles and counter-current operation were proven to not produce high quality heavy or light product. A novel purge recycle step was introduced and improved the carbon dioxide purity. A carbon dioxide purity of 93.8% was achieved by using a rinse step after pressure equalisation steps, but required a compressor and resulted in a significant reduction in carbon dioxide capture rate.

Most granular laundry detergents are manufactured through spray drying. One drawback of this process is wall build-up, which negatively effects process operation, safety and product quality. Macro and micro-scale observations showed the amount and micro-structure of deposits changed significantly across the dryer. These changes were linked to changes in particle properties during drying. Measurements of deposition ranged from 1 - 10 kgm-2, or 2 - 10% of the total slurry sprayed, depending on location, operating conditions and slurry/powder properties. Wall deposition appeared to be time dependent. Wall deposition was broken down into two critical steps; collision frequency, describing how many and how often particles hit the wall and, collision success rate which describes particle’s behaviour upon contact with the wall. Collision frequency was investigated using Particle Imagine Velocimetry (PIV) to measure both fluid and particle dynamics. Finding both to be time dependent, and to vary with position and operating conditions. To investigate collision success rate, particle physical and mechanical properties were studied, revealing mutual dependence of all properties on both formulation and particle size. Impacting these particles at a range of velocities and angles found that the fraction of particles that broke ranged from 0 - 100% and restitution coefficients from 0.1 - 0.8.

The mixing of multiphase turbulent flow within a column is investigated using Positron-Emission Particle-Tracking (PEPT) to understand the controlling phenomena of convection of heat towards the inside surface of its wall in the measurement of the inside-wall heat transfer coefficient by inverse analysis. This is part of a design options study initiated by Rolls-Royce for a key process in the reclamation of uranium from uranium bearing materials. The column contains water and electrolyte and loose metallic swarf. Compressed air is distributed into the column from a perforated plate and hence the multiphase flow is gravity driven. A dynamic assessment of mixing is provided through the framework of the Reynolds-Avergaed Navier-Stokes (RANS) equation in the evaluation of stresses within the multiphase flow. Inverse analysis of heat conduction across the wall of the column provides measurement of the inside-wall heat transfer coefficient. The mixing of swarf coincides with the convergence of the stationary dynamics of the multiphase turbulent flow measured using PEPT. Tikhonov regularisation affords the resolution of the inside-wall heat transfer coefficient of the inverse analysis. A common global gas hold-up between different flows segregates the controlling phenomena of momentum transfer and convection of heat towards the inside surface of the column wall.

Dry powder inhalers (DPIs) are widely used for the therapy of respiratory and pulmonary diseases. In this study, a coupled discrete element method and computational fluid dynamics (DEM-CFD) is employed to investigate the micromechanics of carrier-based DPIs. The effects of van der Waals forces and electrostatic forces on the mixing process, and the influences of air flow and particle-wall impact on the dispersion process are examined. For the mixing of carrier and active pharmaceutical ingredient (API) particles in a vibrating container, it is found that vibration conditions affect the mixing performance. While there is an optimal mixing condition to maximise the number of API particles attaching to the carrier (i.e. contact number) for van der Waals cases, the contact number decreases with increasing vibration velocity amplitude and frequency for electrostatic force cases. It is also revealed that van der Waals forces (short range) and electrostatic forces (long range) result in different mixing behaviours. For the air flow induced and impact induced dispersion, it is found that the dispersion performance improves with increasing air velocity, impact velocity and impact angle, and reduces with increasing work of adhesion. The dispersion performance can be approximated using the cumulative Weibull distribution function governed by the ratio of air drag force to adhesive force or the ratio of impact energy to adhesion energy.

There is a clear need for the development of novel materials for bone regeneration. There is, however, with the exception of poly-(methyl methacrylate) a lack of mouldable polymeric fillers for irregular bone defects. In this thesis, we investigate the production of a new bone graft replacement combining magnetite nanoparticles with polycaprolactone (PCL) to produce a material that can soften in response to the application of an alternating magnetic field. The magnetite nanoparticles were synthesized by a non-surfactant method and then emulsified with PCL. Particle size analysis using light scattering showed that the size distributions of magnetite nanoparticles were influenced by acid concentration and mixing conditions. In contrast, the size of the PCL particles were not strongly related to the mixing conditions, but was influence by the stabilizer used during emulsification. The magnetite nanoparticles showed superparamagnetic behaviour when analysed using SQUID. In the thermal test, magnetite nanoparticles display smooth curves both in the heating and cooling processes and do not shows significant heat loss. Scanning electron microscopy and chemical analysis showed that the magnetite nanoparticles were evenly distributed through the polymer matrix and could be caused to melt following an appropriate external magnetic stimulus. It was demonstrated that magnetite containing microspheres could successful be used for the encapsulation and delivery of antibiotics at a dose which was sufficient to be lethal to E. coli.

The front-loading washing machine is an integral part of household life for many, with little notice given to the extensive interplay of physical, chemical and thermal phenomena taking place as the clothes inside tumble. With over 250 billion washes carried out annually in Europe alone, it has a staggering environmental impact due to its energy and water usage and also due to the chemicals that are drained to the sewage system at the end of a wash. There are two practical routes for reducing this impact; improving the design of the machine so that it requires less energy or water, or improving detergent formulations so that the required level of cleaning is achieved at lower temperatures and with a reduced input of product. The work presented in this study was initially carried out with the aim of designing better detergent formulations. Along the way, the results also enabled new tools for better washing machine design to be developed. For both routes, the basic requirement was to develop an understanding of the mechanisms which result in mechanical cleaning of textiles, which is the focus of results presented in this study.

The selective oxidation of alcohols is one of the most challenging reactions in green chemistry. While the current chemical industry uses organic and inorganic oxidants to produce carbonyl compounds, it is highly desirable to use a heterogeneous catalyst for the efficient oxidation of alcohols. The present research is focused on increasing the activity and selectivity towards the corresponding carbonyl of the heterogeneous oxidation for alcohols. The low activity of 5 wt.%Pt-1 wt. %Bi/Carbon for the oxidation of 2-octanol was investigated in a 500ml stirred tank reactor. The fast reaction rate drops dramatically from 0.23 M/hr to 0.006 M/hr after 15 minutes reaction time when heptane was used as solvent. Different possible causes such as overoxidation, leaching and poisoning were examined. It was found that the loss of high conversion rate was due to product adsorption and hence, different solvents were investigated. A mixture of 16-18% v/v dioxane in heptane was able to effectively regenerate active sites and allow a constant reaction rate of 0.07M/hr. The effects of temperature and pressure were also studied. Gas-liquid, liquid-solid and internal mass transfer effects were determined experimentally and semi-empirically. Six different Langmuir-Hinshelwood rate models were examined where a modified model based on Schuurman et al. (1992) was found to adequately describe the experimental data. The novel 2.5%Au-2.5%Pd/titania catalyst was investigated for the oxidation of benzyl alcohol. Different parameters such as catalyst oxidation state, pressure and stirrer design were studied to increase the activity and selectivity of the catalyst. The selectivity was shown to be highly dependent on the oxygen concentration. Therefore, the pressure, temperature, catalyst treatment were optimised and the reactor configuration was re-designed to enhance oxygen transport to the catalyst. While the conversion rate was unaffected by the oxygen concentration, the catalyst pre-treatment significantly increased the reaction rate. Eventually, the use of a Rushton Turbine at 20 rps with a shower disc sparger and a treated catalyst allowed the selectivity to reach 93%. The reaction could be described with a Power Law model satisfactorily. Transition Metal Oxide catalysts such as \(AgO/ SiO_2, Fe_2O_3/SiO_2, CuO/SiO_2\) and \(CuO/Al_2O_3\) were investigated as an alternative to the expensive noble metal based catalyst. However, it was shown that such catalysts are ineffective for the oxidation of different alcohols by studying the effects of different engineering parameters. The maximum conversion reached was 15% with calcined copper oxide catalyst for 1-octanol oxidation.