Digitized by Kansas State University Libraries

Digitized by Kansas State University Libraries

No description available.

The principal objective of the research programme was to develop a technique for monitoring the condition of power-press cutting tools using ultrasonics. The principle of the technique was based upon the reflection of Surface Acoustic Waves (SAWs) from the tooling's cutting edges, and the main aim of the research was to establish the relationship between reflected SAW amplitude and tool edge condition. Three approaches were used to establish the relationship, namely on-line experiments, off-line bench-top experiments, and mathematical modelling. The modelling work involved using Transmission Line Modelling. Three types of wear were observed on the punch: i. The formation of a wear radius on the cutting edge of the punch. ii. The beating back of the bottom face of the punch adjacent to the cutting edge. iii. The formation of an abraded region on the flank of the punch adjacent to the cutting edge. The research has shown that it is possible to monitor the radius that forms on the cutting edge of the punch using the amplitude of SAW pulses reflected from the edge. Also the work has indicated that the rate of radius formation is a function of the clearance between the punch and the die. However, the effects of the change in the cutting edge's condition on the propagation delay of the SAWs, which was also investigated, was not found to be significant. Whilst in principle it should be possible to automatically monitor the condition and predict the wearing out of press-tools, the development of a working tool-moni toring system capable of monitoring the complete cutting edge of a press-tool is presently being hindered due to the lack of a suitable transducer.

A five degrees of freedom analysis of a perfect precision grinding spindle supported by a pair of back to back angular contact ball bearings is performed. The ball to race contacts are simulated by a non-linear contact spring, representing the elastic deformation of the mating rolling members. Major frequencies associated with various degrees of freedom are observed and a number of design curves, suggesting the best zones of operation for the simulated spindle under radial/ axial loading are also presented. The gyroscopic contribution of an ideal precision spindle was found to be insignificant. The model was further expanded to study the response characteristics of the spindle under lubricated contact conditions. A regression formula is used to model the non-linear spring/ damper arrangement,corresponding to the contact elastohydrodynamic oil film thickness. It is noted that the presence of the oil film along the line of contacts do not significantly alter the position of the major modes of the system. However, its contribution in damping the amplitude of oscillation are found to be significant. Various graphs indicating the overall system response, subjected to varying oil film viscosity, number of balls and the spindle mass are also presented. Furthermore, experimental investigations are conducted to validate the employed methodology. Good agreement is observed between the results of the simulation and the experimental spectra for the fundamental modes of response. Although manufacturing anamolies are not simulated,the formulated models incorporate sufficient versatility to forsee various spindle/bearing configurations, different loading arrangement as well as various geometrical features of a system to be modelled.

This thesis has been motivated by the need to construct reliable, experimentally callibrated models of the drill-string assembly, that capture important phenomena associated with the drilling process, such as torsional oscillations, stick-slip and whirling. This can be achieved using a continuous finite element model, as well as low-dimensional torsional pendulum models in limited cases. The importance of this work lies in the fact, that the experimental rig utilizes real industrial drillbits and rock samples, which after careful identification of TOB speed curves, allows to use an equivalent friction model to accommodate for both frictional and cutting components of the bit-rock interactions. As a first step, a preliminary calibration of the FE model for a straight configuration of the drill-string is carried out, that allows to utilize the low-dimensional model to replicate behaviour of the system. In both of these cases, the agreement between experiments and modelling is achieved. Based on these preliminary studies, it is possible to verify the model in a prebuckled drill-string configuration, for which case a qualitative as well as quantitative agreement with the experiment is obtained. This allows to identify regions where the most dangerous phenomenon of stick-slip is present in parameter space of WOB and top angular speed. Next, the analysis is focused on the whirling phenomenon and calibration of the reduced order rotor model to replicate the motion of the BHA in the bore-hole. A qualitative agreement with the experiments is obtained. An interesting phenomenon of co-existing forward and backward periodic whirling solutions are observed both experimentally and numerically. The influence of the initial conditions and the friction between the BHA and the borehole are investigated, what revealed that there is a threshold value of friction coefficient for which the whirling direction changes from forward to backward.

The flow entering a high-pressure turbine in a gas turbine engine is characterised by a loss of symmetry due to temperature distortions in both radial and circumferential directions, known as hot streaks. In industrial simulations it is common practice to assume uniform inlet temperature conditions to simplify the aerodynamic analysis. However, hot streaks may have significant impact on the turbine aerodynamics with the redistribution of the hot fluid affecting the development of secondary flows with consequent effects on enhanced local heat transfer and aerodynamic losses. The loss of symmetry has also been linked to the excitation of low-order nodal diameter assembly modes of the downstream rotor blades leading to potential blade failure and thus, should be taken into account during the design process. In today’s carbon-constraint environment additional parameters arise as gas turbines are challenged to adapt to variations of the fuel composition driven by the need of efficient and lowCO2 power generation. Introducing syngas, a synthesis gas fuel that is used to power integrated gasification combined cycle (IGCC) power plants, is likely to affect the operating conditions of existing gas turbines leading to the requirement of re-design of components. With particular focus on the turbine hot flow path, the propagation mechanism of hot streaks throughout the turbine will be affected with consequent impact on the turbine aerodynamics and forced response excitation levels originating from the different hot flow patterns. Motivated by the lack of relevant studies, the current work provides a first step towards the evaluation of the effects of syngas on hot streaks aerodynamics and the induced forced response excitation levels. Using full annulus multi-bladerow unsteady 3D CFD simulations and applying combustor representative hot streak profiles in two different gas turbines, a complete analysis of the hot streaks migration is achieved, with respect to a number of geometric parameters such as the hot streaks shape and injection location in both spanwise and circumferential directions, the coolant configurations as well as the combined effects on the secondary flow development. The aerodynamic analysis indicated the propagation of the hot streaks up to the exit of the turbines under investigation with differences in characteristics depending on design parameters. With respect to the effect of fuel composition variations on the blades temperature levels and the flow pattern is observed between the natural gas and syngas turbine with the syngas showing a more concentrated wake shape. In effect of the syngas different flow pattern, differences are observed in the secondary flows with consequent interaction with the hot streaks. Contrast to initial expectations, the forced response analysis iii resulted slightly lower amplitude unsteady force of lower harmonics for syngas compared to natural gas; however, both fuels showed significant levels of the hot streak induced low engine order excitation compared to the burners and stator related blade passing frequency vibration.

Over recent years, with growing concern over climate change, the need for energy which is sustainable, economical and in line with legalisation has led to a substantial surge of interest in organic Rankine cycles (ORC). With the ability to convert low temperature heat sources into power, ORC technology is at the forefront of many sustainable technologies such as biomass, solar, geothermal and waste heat recovery. Despite successful commercialisation for large-scale systems (> 200 kWe), more development is required at the small-scale to realise its potential. For low temperature (< 150 °C), low power applications, volumetric expanders are the preferred choice. However, for a 10 kWe system, a well-designed radial inflow turbine could achieve a higher efficiency, and bridge an observed gap between the output powers of existing volumetric expander systems. This thesis investigates the design and analysis of radial inflow turbines for this application. A thermodynamic ORC model is first developed, which combines cycle analysis with component design. This model is coupled with a multi-objective optimisation, and a novel objective function is developed that considers the trade-off between system performance and system complexity. Following a cycle analysis case study, a radial inflow turbine design method for ORC turbines is developed which extends existing ideal gas design methods to be applicable for real gases. Two candidate turbine designs are developed and are validated using computational fluid dynamics (CFD). For small-scale systems to be economically feasible it is reasonable to assume that the same turbine will be implemented within a number of different systems. This requires off-design models, and the suitability of using non-dimensional performance maps, obtained using similitude theory, has been investigated using further CFD studies. This has led to the development of a modified similitude theory, suitable for subsonic ORC turbines. This modified similitude theory has been implemented within another thermodynamic model, and the results from a case study show how the same turbine can be effectively utilised within a number of different ORC systems. This is done by selecting a working fluid to match the available heat source. Overall, this thesis successfully demonstrates the development of modelling methods for small-scale low temperature ORCs utilising radial inflow turbines. This has considered design and off-design performance models, and ultimately the results demonstrate how the economy of scale of these systems can be improved, aiding in the future commercialisation of the technology.

The construction industry has been challenged by the UK Construction Foresight Panel to apply advanced information and communication technology to improve the performance, in terms of sustainability, of the existing built environment and infrastructure. Traditionally, built-environment maintenance is a capital-cost-driven activity that relies either upon the subjective assessment of a built environment and infrastructure condition (i.e. a stock condition survey) to identify maintenance needs, or upon a reactive response to a component failure. The effectiveness and efficiency of the stock condition survey process to support planned maintenance has previously been questioned and a more sustainable approach, based on an objective assessment of a built environment and infrastructure performance, has been suggested. Previous attempts to develop objective-based (though not performance–based) maintenance models have largely failed, due to the limitations of technology, the daunting task of managing large amounts of data, and the inability of mathematically based models to cope with the complexity of real-life situations. This thesis addresses this challenge by exploring the feasibility of a performance-based assessment methodology to determine the maintenance needs of a buried oil steel-pipeline system and the impact that any changes in condition may have on the performance and integrity of related components in the pipeline system. The thesis also contains an evaluation of the ability and effectiveness of piezoelectric elements in pipeline defect (crack) signature detection to predict changes in component performance with data sets derived experimentally using laboratory bench testing. Vibration sound-emission detection techniques performed on various oil steel-pipeline defects, using non-destructive testing methods, were validated using attenuation and waveform analysis. Defect size and progression (i.e. the pattern characteristics of the defect) were monitored, measured and identified through spectrum analysis of multiple emission signals in combination with a number of frequency bands. Two series of tests were undertaken to evaluate the ability of vibration sound emission characteristics to identify steel pipeline defects, including leakage. Test Series 1 established the frequency (waveforms) of the generation of the acoustic emission signal caused by normal fluid dynamics (water flow) through the experimental steel pipe and the resulting signal propagation characteristics. Test Series 2 detected and monitored changes in the signal characteristics for incipient defects: (a) small-nail damage, (b) medium-sized nail damage, (c) large-nail damage and (d) crack to leakage source [sealed holes as a simulated corrosion to total failure]; oil was the fluid medium. The defect sources and leakage signals were also studied, and compared with theoretical models. The results of the theoretical analysis and the laboratory experiments confirmed the ability of non-destructive testing, based on vibration sound emission techniques, to detect and distinguish between different failure modes. The ability to carry out a basic inspection, analysis and report of a pipeline using an integrated-sensor device offers many potential benefits. The use of an integrated-sensor device is expected to provide valuable pipeline management information. Specifically the ability to detect and locate mechanical damage at the incipient stage and provide an assessment of the overall pipeline operating condition, including changes in performance profile and prediction of an estimated time to failure, has been shown to be feasible as part of a pipeline maintenance and rehabilitation programme.

published_or_final_version / Business Administration / Master / Master of Business Administration