This thesis describes research undertaken to improve a technique for the measurement of the thermal conductivity of molten materials. The research follows on from the work of previous researchers who designed and tested an instrument for the measurements of the thermal conductivity of molten metals up to 750 K. The previously used transient hot-wire technique, which consisted of the experimental measurement of the voltage response of a sensor and a subsequent inverse Unite element analysis, has been significantly upgraded. The experimental part of the technique has been improved by the introduction of a new design of the sensor for the measurement of the thermal conductivity. Both the new and the original designs have been used to investigate the same material samples in order to demonstrate the robustness and repeatability of the experimental technique. Additionally, the finite element analysis employed has also undergone various major improvements and resulted in a new finite element model which not only represents the true geometry of the experimental device but also employs a more accurate solution of the transient, conductive heat transfer. The significant upgrade of the technique and the availability of two different sensor designs have helped to uncover systematic errors which could not have been previously identified and may have resulted in deviations of the measured thermal conductivity. Five original sensors and five sensors with the new design have been used to investigate the thermal conductivity of molten indium, tin and lead at various temperatures up to 750 K. The results have been compared to previously published data and the discrepancies have been discussed and explained. Each metal has been measured using at least two sensors and the consistency of the measured data has also been verified by using two different samples of pure tin. Besides the pure metals, the thermal conductivity of several metal alloys currently used in industry has been investigated within the same temperature range. The overall uncertainty of the measurements of the thermal conductivity is estimated to be ±3 %.

The focus of the present work is to develop stochastic reduced basis methods (SRBMs) for solving partial differential equations (PDEs) defined on random domains and nonlinear stochastic PDEs (SPDEs). SRBMs have been extended in the following directions: Firstly, an h-refinement strategy referred to as Multi-Element-SRBMs (ME-SRBMs) is developed for local refinement of the solution process. The random space is decomposed into subdomains where SRBMs are employed in each subdomain resulting in local response statistics. These local statistics are subsequently assimilated to compute the global statistics. Two types of preconditioning strategies namely global and local preconditioning strategies are discussed due to their merits such as degree of parallelizability and better convergence trends. The improved accuracy and convergence trends of ME-SRBMs are demonstrated by numerical investigation of stochastic steady state elasticity and stochastic heat transfer applications. The second extension involves the development of a computational approach employing SRBMs for solving linear elliptic PDEs defined on random domains. The key idea is to carry out spatial discretization of the governing equations using finite element (FE) methods and mesh deformation strategies. This results in a linear random algebraic system of equations whose coefficients of expansion can be computed nonintrusively either at the element or the global level. SRBMs are subsequently applied to the linear random algebraic system of equations to obtain the response statistics. We establish conditions that the input uncertainty model must satisfy to ensure the well-posedness of the problem. The proposed formulation is demonstrated on two and three dimensional model problems with uncertain boundaries undergoing steady state heat transfer. A large scale study involving a three-dimensional gas turbine model with uncertain boundary, has been presented in this context. Finally, a numerical scheme that combines SRBMs with the Picard iteration scheme is proposed for solving nonlinear SPDEs. The governing equations are linearized using the response process from the previous iteration and spatially discretized. The resulting linear random algebraic system of equations are solved to obtain the new response process which acts as a guess for the next iteration. These steps of linearization, spatial discretization, solving the system of equations and updating the current guess are repeated until the desired accuracy is achieved. The effectiveness and the limitations of the formulation are demonstrated employing numerical studies in nonlinear heat transfer and the one-dimensional Burger’s equation.

Rankine source method coupled with Mixed Euler-Lagrange (MEL) algorithm is developed to investigate wave-body problems. Under Euler specification a boundary-value problem is solved by placing fundamental singularities outside the computational domain and satisfying the boundary conditions at prescribed control points. At every time step, Lagrangian frame is applied to update the control points position during regridding process. A space increment method for source points distribution incorporating horizontal free surface source arrangement and vertical desingularized distance is developed and this method connects free surface panel to body panel size. By reducing the number of source points, this method significantly increases the computational efficiency. A single node scheme is implemented to treat intersection points. This scheme regards intersection points only as body panel ending points. The first source points on the free surface are placed away from the intersection points and generated wave is started from these source points rather than intersection points. During regridding process, body panel number keeps constant and panel size varies to match the variation of wetted body surface. In the process of repanelling the free surface, panels slide horizontally due to the variation of wetted body surface pushing them back and forth. After their horizontal positions are fixed, the source points follow the wave elevation and are located on the updated wave surface in the vertical direction. A least square based smoothing technique is developed to eliminate the "sawtooth" phenomenon occurred in the free surface updating for two-dimensional fully nonlinear problem. Both two- and three-dimensional forced body oscillatory motion problems are studied and extensive comparisons show a good agreement with published results. The methods developed are proved to be accurate, efficient and robust for wave-body problems.

The work described in this thesis is concerned with the performance of the mechanical system of a mobile robot that is capable of omnidirectional motion. The main attribute of such mobile robots is that their direction of motion is independent of chassis orientation. This attribute endows them with exceptional manoeuvrability, but it is also found to pose substantial problems by changing the level of accuracy and stability of the robot as its direction of travel changes. The main objective of the research is to conduct a detailed evaluation of the performance of a mobile robot which is capable of omnidirectional movement achieved by means of a synchronized all-wheel steering and all-wheel drive (Synchro-drive) technique. The objective is met by comparing the synchro-drive method with other configurations used for mobile robots, by comparing different designs of the synchro-drive method and by analyzing synchro-drive mechanical behaviour in response to drive and steering inputs. A kinematic model of the synchro-drive arrangement is formulated and this is used to analyze different designs and to assess the limits of the control variables beyond which a Synchro-Drive Mobile Robot (SDMR) operation will become unstable. A new version of the synchro-drive arrangement was developed and was used to perform extensive practical testing in order to determine factors affecting positional accuracy and the trajectory actually executed by the mobile robot. The analysis of the boundaries of the control space revealed the limits on acceleration which may be allowed by the robot's control system for it to remain stable. It also showed that the acceleration limits depend on the angle between the wheel heading and the chassis orientation, which is defined as the robot's posture. Practical experimentation identified the major influences on robot accuracy and also related the form, magnitude and direction of these errors to the robot's posture. The experiments revealed that the errors were due partly to aspects of the design itself and partly due to inevitable errors in the complete mechanical system. A continuous position error correction method is proposed which uses experimental data as the basis for correction. Correction quantities vary with posture, and the method uses a modification to the steering rate to minimize trajectory error. Overall the study reveals the factors which must be considered to enable the potential of the synchro-drive mobile robot to be fully realized.

Kansei or affective engineering is the discipline of designing products to be psychophysically more appealing to the human mind and senses. Touch-feel perception of the materials used in consumer products ranging from portable electronics, furniture to automotive interiors plays an important role in the attractiveness of a product. Touch-feel perception is a qualitative measure and is an extrinsic property of the material. To better assist designers and material scientists to optimise aspects of a material for touch-feel perception, it is important to find a link between the qualitative touch-feel attributes with quantitative intrinsic properties of the materials. There is ongoing research in trying to decipher the links between touch-feel perception expressed through semantic psychophysical descriptor words, to physical parameters of the material sample such as the surface topographical, mechanical and tribological attributes. The objective of this work is to fill the current knowledge gap between micro-surface physical properties and customer's perceptual response to surface tactile sensory information as well as their affective preference through theory, correlation models and experimentation. A conceptual framework of surface tactile evaluation system can be divided into three parts: measurement of the surface physical characteristics, sensory evaluation and correlation analysis. To this end, the thesis documents the development of a friction measurement apparatus including an artificial finger to estimate the friction of a material against human skin in an accurate and repeatable manner. Secondly, correlation analyses were performed on the skin-against-material friction and the tribological factors, including the material surface parameters (e.g. roughness) and physical characteristics (e.g. hardness) of various metal and thermoplastic materials. Finally, the human touch-feel perception was assessed through a questionnaire and the results were modelled to obtain a link between the tribological factors and touch-feel perception. Generally, human beings feel a surface by stroking or sliding one's finger, which experiences friction. It is challenging to objectively describe the friction experienced by a human finger with respect to surfaces being stroked, as different surfaces and different working conditions can all influence the results. In order to understand the interaction between different surfaces and the friction experienced by a human finger, one has to minimise the variation due to human fingertips and touch conditions across experiments, such as fingertip humidity, temperature and elastic properties. To achieve this, a friction measurement apparatus incorporating an artificial fingertip has been developed. The artificial fingertip is made of multi-layered materials to mimic the structure, shape, softness and friction properties of a real human fingertip. The friction test apparatus consists of the artificial fingertip, a linear flexure mechanism and a reciprocal linear stage. It is capable of measuring the contact force and friction force simultaneously to give an estimate of the friction coefficient of the material-under-test. Twelve aluminium samples and five steel samples of different surface finishes were tested under different contact forces and stroking speeds. Comparisons were made between the friction results measured in vivo by a human fingertip and those by the artificial fingertip. The results have shown that for the material samples investigated, measurements from the artificial finger achieved a high correlation with results from real human fingers (r2 = 0:8 ~ 0:98) for surface ground steel and milled aluminium. Therefore the artificial finger can be used to mimic the friction characteristics of a real human fingertip and more importantly to measure the skin-against-material friction accurately and in a repeatable way. In addition, in order to better understand the contact mechanism between the artificial finger and the surface, a suitable theoretical model which incorporates how the contact force relates to the contact area is essential. To enable the modelling of the contact mechanism, the Young's modulus of the artificial fingertip has to be identified, as it is an essential input parameter for all contact theory models as well as FEM. The artificial finger was measured by using micro- and nano-indentation with Berkovich/spherical-tipped indenters. The contact area measurement was conducted by loading a custom-built glass plate on the artificial fingertip and observing the contact area under an optical microscope. Hertz theory was used to model the fingertip and predictions were compared against finite element analysis. The results support the fact that the Hertz contact theory is valid for modelling the contact mechanism of the artificial finger. Thermoplastic elastomers (TPE) and copolymers of elastomer are commonly used in manufacturing car interiors to give the surface a less harsh and more pleasing feel. Ongoing research has been trying to decipher the links between touch-feel perception expressed through semantic psychophysical descriptor words, to physical parameters of the material sample such as the surface topographical, mechanical and tribological properties. A series of five patterned and five coated TPE surfaces provided by an automotive manufacturer were characterized-topographical parameters of the samples by a surface profiler and mechanical/tribological parameters by a nanoindenter. The friction characteristics of these specimens were measured by the friction test apparatus and the artificial finger. The results showed that the artificial finger is representative of a human finger in its friction-sensing capability. In the second part of the thesis, the relationship between the skin-against-material friction coefficient and the surface topography parameters Rq and Sm were deduced according to Hertz contact theory. The theory gives good agreement with experimental results. In addition, the relationship between the friction coefficient and the other mechanical parameters such as the Young's modulus, skewness, kurtosis, surface slope were investigated through correlation analysis. Finally, 54 people of different age and gender were asked to rank the specimens in terms of 5 pairs of psychophysical descriptors, such as `rough/smooth', `cold/warm', `slippery/sticky', `soft/hard' and `like/dislike'. A rank-ordered logit model was deployed to correlate the human touch feel perception rankings and the thermoplastic samples, and the results were compared with correlation methods used in previous work. The results indicated the specific parameters which are correlated with human touch-feel perception and also their relative contributions. The results form a good guideline for material scientists and designers to, for example, build more touch-desirable car interior materials and consumer packaging.

Due to growing concerns about climate change, the heat and power sector is continuously facing challenges to reduce CO2 emissions. Carbon capture and storage (CCS) is one of the short-medium term measures that can mitigate CO2 emissions emitted from fossil fuels utilisation. Oxy-fuel combustion is a promising technology for CSS that can be integrated into the new and the current fleet of power plants. Biomass is a carbon neutral renewable source of energy that can replace fossil fuels. If the biomass is utilised as a fuel in oxy-fuel combustion it could lead even to negative CO2 emissions. Although the sintering and agglomeration problems associated with the combustion of non-woody biomasses in the fluidised beds are still major issues, fluidised beds have emerged as one of the best among the other proven biomass combustion technologies, mainly due to their fuel flexibility, low SOx and NOx emissions. However, oxy-fuel combustion technology in fluidised beds is in the early stages of development and still needs a lot of research for improvement before its application on full-scale power plants. In this work basic combustion fundamentals of different biomass fuels in terms of energy production were studied using thermogravimetric analysis (TGA) under air, N2, CO2 and selected oxy-fuel (30%O2/70%CO2) reaction environments. Then a 20 kWth bubbling fluidised bed combustor (BFBC) was designed, manufactured and successfully tested for a range of biomass fuels under air and oxy-fuel combustion environments. The agglomeration and sintering behaviour of these biomass fuels during combustion under air was also investigated using different analytical techniques such as SEM-EDX, XRD and XRF. The biomass fuels investigated in this study include domestic wood, industrial wood, miscanthus, wheat straw and peanut shell pellets. The BFBC testing of these biomass fuels focused on the influence of operating conditions, the effect of excess air level and fuel feed rate on the hydrodynamics, temperature profiles and emissions, NOx, CO2 and CO within the BFBC. Air staging can be very effective in reducing NOx emissions of non-woody biomass fuels especially when the secondary air was injected at the higher level with an overall low excess air level. A maximum NOx reduction percentage of 30% was achieved for the non-woody biomasses during air staging combustion. The non-isothermal TGA analyses under N2 and CO2 showed almost identical weight loss (R), reactivity (RM) and activation energy (Ea) profiles in devolatilisation zones. However, when devolatilisation occurred under CO2 conditions at temperatures higher than 700 oC, an additional weight loss was observed for all biomass fuels, being indicative of the contribution of CO2-char gasification reactions. Under air and oxy-fuel (30%O2/70%CO2) results showed almost similar profiles for R, RM and Ea. In oxy-fuel atmospheres, by replacing N2 with CO2 a slight increase in the maximum rate of weight loss (RMax) was observed in both reaction zones for all studied biomasses. The unstaged and staged air combustion experiments in the 20 kWth BFBC showed that higher excess air always led to higher NOx emissions for any of the biomass fuels tested because less CO and char were available in the reactor to promote NOx reductions. Due to the consequence of the high volatile matter content of the biomass fuels, the maximum temperatures were achieved at the top of the dense bed and/or beginning of the freeboard, which suggests that the main combustion reaction takes place in this part of the combustor. Air staging leads to higher temperatures in the freeboard, especially at low excess air levels, as a result of additional combustion in the freeboard under staged air conditions. Air staging can be very effective in reducing NOx emissions of non-woody biomass fuels especially when the secondary air was injected at the higher level with an overall low excess air level. A higher percentage of carbon in ash was obtained while working under air staging conditions than that of without air staging combustion. The results of oxy-fuel combustion tests in the 20 kWth BFBC showed that oxy-fuel combustion was different from air combustion in several ways, including reduced gas temperatures, delayed flame ignition, increased CO emissions under 21%O2/79%CO2 and 25%O2/75%CO2 oxy-mixtures. Many of these parameters were associated with differences in properties of the main diluting gases CO2 and N2 in oxy-fuel and air combustion respectively. In order to match the biomass oxy-fuel combustion gas temperatures to those of air combustion, the oxygen concentration in the mixture of O2/CO2 has to be increased to 30% or higher. Moreover, the oxy-fuel combustion with 30%O2/70%CO2 has shown higher efficiencies than air, which indicates biomass fuels can be successfully combusted in the BFBC under oxy-fuel combustion conditions. The agglomeration and sintering behaviour was observed under continuous air combustion conditions in the 20 kWth BFBC. The analysis using SEM-EDX, XRD and XRF concluded that the potassium present in wheat straw was mainly responsible for agglomeration, which was detected in the form of KCl and K2O in the bed material and cyclone ash samples.

Gas flow in columns of high viscosity liquids is found in heavy oil and bitumen production, polymer manufacturing and in Volcanology as silicate magmas in the volcanic conduit. Predicting the characteristics of the hydrodynamics of gas flow, under such conditions is essential in both design and safety assessments. In oil and gas industry, it is very important in the design of the equipment and the very long pipelines. Further, it is important in the design and safety of the industrial equipment and the pipelines in the polymer manufacturing. Finally, the ability to predict natural phenomenon and develop the knowledge about volcanoes activity and the nature of eruptions in volcanoes is important for the assessment of environmental risks. The majority of the works, which have studied gas-liquid flow in pipes, have been carried out mainly by using water or liquids of low viscosities (< 1 Pa.s). Knowledge, regarding the hydrodynamics of gas flow in high viscosity liquids and large diameter columns, is still limited though despite the importance of this subject. In this work, the characteristics of gas-high viscosity oils in large diameter columns were studied over a wide range of gas flow rates. Two column geometries were used: one of 240 mm and the other of 290 mm internal diameter. The columns were initially filled with stagnant Silicone oil of viscosities 360 and 330 Pa.s respectively. Electrical Capacitance Tomography (ECT) technique was employed for the data measurements besides a high-resolution camera. In general, 4 flow patterns are found over the selected range of gas flow rates. These are seen to differ from the ones in lower viscosity liquids. First, bubbly flow consists of single spherical bubbles that rise at a constant velocity in the centre of the column at low gas flow rates. Second, slug flow which consists of long bubbles (Taylor bubble) with rounded top and end, with a diameter almost equal to the column diameter and separated by liquid slugs. The third flow pattern is the transition to churn flow which occurs due to further increase of gas flow rate. The falling film, around the very long bubbles, accumulates to create regions of high frequency activity that consists of liquid bridges. These regions are named “churn” regions. The length of this region increases gradually with increasing the gas flow rate. At very high gas flow rates, the gas flows through an open wavy non-symmetrical core in the column of the viscous oil (churn regions). The length of the churn regions increases significantly at this flow regime which can be named churn flow regime. Small bubbles of millimetres to centimetres diameter are seen to accumulate in the column due to the high viscosity and low velocity of the liquid motion. These bubbles generate due to the bubble eruption at the top section, bubble coalescence along the column and at the gas injection points at the bottom of the column.

The energy conversion process of fossil fuels to electrical power is accomplished using inefficient technologies with negative consequences on human health and environment as well as contributing to depletion of finite natural resources. Therefore, there is currently increased societal pressure to adopt more environmentally sustainable and low carbon solutions. The share of energy derived from renewable sources (solar, wind, biomass, etc.) is projected to increase significantly over the next decades to decarbonise the energy sector. However, to achieve this objective, the heat conversion technologies need to make a step improvement in energy conversion. Stirling engine is considered to operate on the most efficient heat cycle capable of operating on a range of fuels including renewables. This research investigated a novel design of a diaphragm Free-piston Stirling Engine for low temperature power generation in remote and inaccessible regions of the world. The design incorporates a flexible edge welded bellows material to support the displacer and a flat elastomer as a power piston. This mechanical arrangement of the moving parts of the engine eliminates air leakage, mechanical friction of the power piston and cylinder, and low spring losses. A proof-of-concept and a validated mathematical model was developed as part of this project. The mathematical model was based on solving the energy, mass and momentum conservation equations of the working fluid in different parts of the engine. The performance of the engine was evaluated for different design parameters such as temperature, pressure, operating frequency, etc. A proof-of-concept prototype was built and tested under controlled laboratory settings to measure the energy performance. It was demonstrated the proof-of-concept engine can operate successfully at low to medium temperatures (up to 300oC) at atmospheric pressure and frequency of 16Hz. The tests also showed that under sufficient temperature gradient the engine is self-starting. Though the shaft power output was insignificant for the size of the engine, the design and laboratory results have contributed to advancing the technology and its application.

Space exploration missions often use drills or penetrators to access the subsurface of planetary bodies. Protected by the conditions experienced at the surface, these regions have potentially been untouched for millennia. As such, the subsurface is a very attractive option for scientific goals, be it the search for extra-terrestrial life, to examine the history of the planet, or to utilise underground resources. However, many issues arise in such a task. Every other rocky body in our solar system possesses a surface gravity lower than our own, resulting in a lower available weight for a spacecraft to ‘push’ on a penetrating device. Add to this the low power availability and complications regarding remote operation, and this becomes a very difficult process to achieve. Mole devices which burrow through the ground whilst tethered to a surface-station to provide power and data have shown great promise in this regard. Using an internal mass to ‘hammer’ themselves into the ground, special care is required to ensure internal components are not damaged, and that they can arrive at their target depth in a reasonable period of time. There is continuous development in these types of drilling and penetrating technologies and anything that can penetrate with a lower weight-on-bit (WOB), and consume less power, could potentially be extremely useful for these situations. High powered ultrasonic vibrations have been shown to reduce operational space and forces required in cutting bones for surgery. Additionally, they have been successful in reducing WOB requirements for drilling devices through rocky substrates. To maximise penetration depth, it is often favourable to progress though granular material rather than solid rock, however this also provides its own set of problems. This work looks at applying ultrasonic vibration to penetrating probes for use in granular material, with the aim of utilising it in low gravity or low mass scenarios. Before this can be done however, the regolith used for testing must be fully characterised and consistent preparation methods established, ensuring that all other effects are accounted for. An ultrasonically tuned penetrator was designed and manufactured, and the effects it had on the surface of sand were investigated using a high-speed camera and optical microscope. It was found that regions of sand immediately surrounding the penetrator were highly fluidised, localising any deformations to a small radial distance. Penetration tests were then conducted that showed ultrasonic vibration significantly reduces the penetration forces and therefore the overhead weight required, in some cases by over an order of magnitude. A similar effect was seen in power consumption, with some instances displaying a lowered total power draw of the whole system. Experiments were then conducted in a large centrifuge to examine the trends with respect to gravity. Gravitational levels up to 10 g were tested, and the general trend showed that ultrasonic penetration efficiency indeed increased at lower gravities, suggesting that the force reduction properties would be enhanced at lower levels of g. Finally, the first steps to applying this technique as a fully-fledged penetration device were conducted. These tests oversaw combining ultrasonic vibration with the established hammering mechanism used by mole devices. Comparing this against a purely hammering penetration, it was found that the addition of ultrasonic improved performance significantly, greatly reducing the number of strikes required to reach the same penetration depth. To conclude, the work presented in this thesis shows the potential that ultrasonic vibration can have with advancing low gravity/low mass penetrating devices. Reducing both the weight and power requirements can be a huge boon to small spacecraft, and the potential use as subsurface access or anchoring devices makes it an attractive avenue for future research and development.

We investigate multiple crack evolution under quasi-static conditions in an isotropic linear-elastic solid based on the principle of minimum total energy, i.e. the sum of the potential and fracture energies, which stems directly from the Griffith’s theory of cracks. The technique, which has been implemented within the extended finite element method, enables minimisation of the total energy of the mechanical system with respect to the crack extension directions. This is achieved by finding the orientations of the discrete crack-tip extensions that yield vanishing rotational energy release rates about their roots. In addition, the proposed energy minimisation technique can be used to resolve competing crack growth problems. Comparisons of the fracture paths obtained by the maximum tension (hoop-stress) criterion and the energy minimisation approach via a multitude of numerical case studies show that both criteria converge to virtually the same fracture solutions albeit from opposite directions. In other words, it is found that the converged fracture path lies in between those obtained by each criterion on coarser numerical discretisations. Upon further investigation of the energy minimisation approach within the discrete framework, a modified crack growth direction criterion is proposed that assumes the average direction of the directions obtained by the maximum hoop stress and the minimum energy criteria. The numerical results show significant improvements in accuracy (especially on coarse discretisations) and convergence rates of the fracture paths. The XFEM implementation is subsequently applied to model an industry relevant problem of silicon wafer cutting based on the physical process of Smart-CutTM technology where wafer splitting is the result of the coalescence of multiple pressure-driven micro-crack growth within a narrow layer of the prevailing micro-crack distribution. A parametric study is carried out to assess the influence of some of the Smart-CutTM process parameters on the post-split fracture surface roughness. The parameters that have been investigated, include: mean depth of micro-crack distribution, distribution of micro-cracks about the mean depth, damage (isotropic) in the region of micro-crack distribution, and the influence of the depth of the buried-oxide layer (a layer of reduced stiffness) beneath the micro-crack distribution. Numerical results agree acceptably well with experimental observations.