Gas turbine emissions are indicated to be an environmental and local air quality problem. NOx, SOx and CO2 are linked with various greenhouse phenomena, and particulate matter (PM) is a local air quality issue. The measurement of PM is currently based on Smoke Number (SN), an outdated reflectance scale. A dual fuel Dry Low Emissions (DLE) combustor has been tested, with a focus on qualifying NOx emission processes. An alternative fuel (diesel) was used to generate a 3D data-set under gas turbine-relevant conditions, enabling the confirmation of CFD models for liquid-fuel combustion. Sampling line configurations were investigated to provide the aircraft engine manufacturing industry with a modern methodology of quantifying PM. Observations between different flow regimes and temperatures with PM size distributions show significant differences. The Differential Mobility Spectrometer (DMS500), a fast mobility sizer, was commissioned on an atmospheric pressure rig, using aviation-grade kerosene, to define the PM sampling methodology. On legacy combustors and a modern lean burn combustor, PM size, number and mass from the DMS500 were compared to other instruments and methodologies (i.e. TEM, CPCs, AD-MAAP) showing good correlation. Poor agreement was noted with gravimetric analysis. Size distribution comparisons between DMS500 and a scanning mobility particle sizer (SMPS) contain significant variations, due to residence time and particle condensation from the SMPS sampling configuration. Generally, engine exhaust aerosol size distribution was observed to be bimodal with a peak at 15-25nm and a peak at around 50-80nm. Finally, a large civil gas turbine was tested using emission certification representative sampling methodology. Real time sizing data showed good reproducibility, identifying both nucleation and accumulation mode PM. Good agreement was found with mass measurement on the first-order approximation (FOA3.0) SN correlation and the DMS500, whereas the MAAP failed to track the engine conditions without a correction. The sample line temperature, residence time and dilution affect the size distribution of the PM measured from engine exhaust aerosol. Quantitative size analysis of PM under controlled sampling conditions shows good reproducibility, though the effects of volatile organic matter losses need to be quantified further. Insight was developed into the formation processes of combustion product species within a DLE combustor.

The research concerns the condition monitoring of cylinder liner of large bore diesel engines using various methodologies to identify the onset of scuffing. The reasons of scuffing, improved designs and operational processes to prevent its occurrence were discussed. The research focused on modeling the normal condition of the cylinder liner with sufficient lubrication and detecting the precursor of scuffing by reducing the lubrication. The four detection systems used on the test facilities and field tests of the cylinder liners used tribology, temperature sensor, eddy current sensor and acoustic emission sensor. Experimental assessment of eddy current sensor was conducted for insufficient lubricating oil conditions for different cylinder liner wall pressures using a specially designed test facility. Field tests of temperature sensor and eddy current sensor were carried out on a 800mm bore worn cylinder liner of a container ship in service. Field test of acoustic emission sensor was carried out on a high speed automobile engine. Scuffing detection by temperature sensing should be considered as the last safety barrier, as it registers the after effect of scuffing and solely depends on the localized condition and the material’s thermal status. Mounting eddy current sensors are considered intrusive. Four sensors per cylinder are needed, which are prone to damage by the rings when the liner wears out. Additionally, the sensors measure only small section of the rings and their lubrication condition. Acoustic emission analysis effectively detects onset of scuffing on the cylinder liners and the rings. Initial findings from the lab and the field test on a four stroke engine confirmed this. However, more field tests under various loading condition on a slow speed engine is needed to understand the various event alignment and the non-routine detection, such as scuffing. They can be coupled with thermocouples to serve as a secondary protection.

This thesis focuses on the aerothermal analysis of the intermediate pressure compressor (IPC) drum of a jet engine and the associated internal fluid flow. Emphasis is given to the CFD modelling of the drum internal flow, associated heat transfer and coupled FEM/CFD aerothermal analysis of the drum for a flight cycle. At the engine operating conditions two main types of flow occur in the internal cavities of the IPC drum; radial inflow and axial throughflow. Evaluation of two turbulence models, the k-E model and the Reynolds stress model (RSM), for the prediction of these flows, is one of the aims of the present study. Experiments of radial inflow in a rotating cavity by Firouzian et a1. (1985) and Farthing (1989) showed a high degree of swirl velocity in the core region of the cavity and this rotating core is found to influence the wall torque and heat transfer to a great extent in such cavity flows. Various test cases are considered to evaluate the turbulence models for the prediction of radial inflow in narrow and wide stationary cavities, rotating cavities with different radius ratios alb, gap ratios s /b and shroud geometry. The rotating cavity test cases covered a range of rotational Reynolds number (Re</>,b) from 1.7 x 105 to 1.2 X 106 , mass flow parameter (Cw ) from 1,300 to 10,100 and inlet swirl fractions from 0 to 1.0. The RSM performed reasonably well in predicting the flow and heat transfer for all the tested cases. Axial throughflow experiments by Farthing et a1 (1991) showed vortex breakdown of the central jet and a strong correlation of jet breakdown modes to the jet Rossby number Ro. Unsteady Reynolds-averaged Navier-Stokes (URANS) models and large eddy simul~~ions (LES) subgrid scale models are assessed for the prediction of such axial throughflow by carrying out a full 360 degree time dependent simulations. The RSM and LES predicted the flow field and breakdown frequencies reasonably well. Based on the above rotating cavity validation studies and the standalone CFD studies of the IPC drum internal flows, the 2D k-E model and 3D RSM are selected for the coupled FEM/CFD simulations. Predictions from the coupled FEM/CFD simulations are compared with the engine test measurements. These simulations showed significantly improved predictions of the metal temperatures inside the IPC drum at various locations. This work will be a valuable addition to the ongoing efforts to carry out the whole engine computational model at the Surrey UTC and Rolls-Royce pIc

Several technologies offer a promising strategy for the exhaust after-treatment of diesel engines in which continuing to further thrift Pt/Pd alloy diesel oxidation catalysts (DOC) by increasing Pd content, the prospect of an essentially Pt-free DOC catalyst is very attractive because Pd is cheaper than Pt. However, to determine the viability of this approach a much better understanding of PdO/AI20 3 chemistry and reactivity is needed. Therefore, the aim of this work is (1) to improve the catalytic performance of the Pd based ?iesel oxidation catalyst by using different preparation method and to investigate the effect of the thermal aging' on the catalyst activity towards CO and C3H6 oxidation. (2) clarify the effect of H20 and NO in the CO and C3H6 catalytic oxidation reactions in terms of reaction mechanism. (3) to investigate the catalyst deactivation induced by sulphur poisoning. It has been found that the structure, morphology and textures properties of the catalysts were significantly affected by both the preparation method and the thermal aging. In particular, the formation of highly catalytically active palladium nanoparticles showing a high density of defects and/or corners, have been observed when the sample was prepared by solution combustion synthesis and when thermal aging was carried out between 500 - 750 ·C. Regardless of the preparation method and the thermal aging, the presence of water in the reaction mixture resulted to promote the catalyst performances with the respect to CO oxidation which occurred via reaction with water derived species rather than dioxygen. Beside, water has also been found beneficial in modulating the NO, C3H6 and S02 inhibition. The formation of strongly bonded sulphate species on the catalyst surface along with the permanent loos of the structure and morphology properties of the catalyst, have been identified as the main cause of the catalyst deactivation induced by sulphur poisoning.

During sea-level transients, such as engine start-up or shut-down, ramping chamber pressure causes the nozzle exit pressure to fall below ambient conditions, resulting in flow over-expansion. Internal shockwaves form that incite unsteady boundary layer separation producing an asymmetric internal pressure field that manifests as dynamic off-axis loads. These side-loads reduce the safe-life of the vehicle and have also be known to cause sudden catastrophic failure. As a result, rocket nozzle area ratio is purposely limited to ensure that flow separation does not occur elsewhere in the mission profile and as such, the vacuum performance of the vehicle is reduced by as much as 20%. An experimental study comparing the side-load distributions of four conical nozzles with wall angles, 8.3°, 10.4°, 12.6° and 14.8° a truncated ideal contour and thrust optimised parabolic nozzle has been carried out. Direct side-load measurements taken using a strain tube have shown that conical nozzle wall angle has very little affect on the side-load magnitude. The truncated ideal contour nozzle displayed the lowest side-load distribution and was found to be approximately 50% lower than the magnitudes produced by the thrust optimised parabolic. An analytical model has been developed to simulate the side-load distribution across geometries which only produce free-shock separation. A universal method of replicating the pressure distribution across a free-shock separated nozzle was first developed, validated with a high level of confidence against three nozzle geometries. This was used in conjunction with a shock excursion model, whereby the oblique shock relation was perturbed to first order in order to generate internal pressure field asymmetry. Comparisons made to experimental results have shown the side-load model can predict distributions with errors as low as 3.58% for truncated ideal contour nozzles.

Two corotating discs and a peripheral shroud can be used to model the cavity created by compressor or turbine discs in a gas-turbine engine. This thesis describes an experimental investigation into the effect of geometry on the flow and heat transfer in a rotating cavity for three cases: radial outflow, radial inflow and axial throughflow. The study included flow visualisation, pressure and heat transfer measurements. The experimental rig comprised two discs of outer radius 381 mm and a peripheral shroud; the axial distance between the discs was varied up to 171 mm. Air entered or left the cavity through either holes, of outer radius 38.1 mm, in the centre of the discs, or through discrete holes in the shroud. The cavity was rotated up to 2000 rev/min with flow rates of up to 0.2 kg/so One disc could be radiantly heated to produce a temperature profile that increased with increasing radius; the maximum front face temperature was 100°C. Flow visualisation, for the axial throughflow case, was also conducted on a second smaller rig. In a rotating cavity with a radial outflow, it was shown that whilst cobs at the inner radii do not significantly affect the flow structure, relative to the plane-disc cavity, protrusions at the outer radii result in the flow becoming nonaxisymmetric; there was also axial flow in the core. Heat transfer measurements, using the transientanalysis technique, for the cavity with cobs but without protrusions, were reasonably predicted by the solutions of the nonlinear integral equations. For the radial inflow case, radial fins attached to one disc or de-swirl nozzles at inlet to the cavity were both shown to be effective in reducing the pressure drop across the cavity. Measured pressure drops agreed reasonably well with the linear and nonlinear solutions of the momentumintegral equations. Comparison between measured values, using the transient analysis technique, and predicted values of the Nusselt number in a cavity with cobs showed reasonable agreement, provided that Ekman-layer flow occurred. In a rotating cavity with an axial throughflow, flow visualisation revealed that the nonisothermal nonaxisymmetric flow structure differed significantly from the axisymmetric flow in an isothermal cavity. Thermally insulating cobs were shown to affect the flow structure, especially when the axial distance between the cobs was small. Heat transfer measurements, using fluxmeters, showed that the Nusselt number increased with both rotational speed and flow rate. The results were also consistent with measurements obtained from other experimental rigs.

Future industrial gas turbines will be required to operate at higher temperatures to increase operating efficiencies and will be subjected to more frequent thermal cycles. The temperatures that the substrates of components exposed in the harshest environments experience can be reduced using air-cooling systems coupled with ceramic thermal barrier coatings (TBCs); however, few studies have been carried out at the substrate temperatures encountered in industrial gas turbines (e.g. 900 – 1000 °C). Better understanding of their behaviour during service and, their various potential failure mechanisms, would allow more accurate prediction of TBC lifetimes and improve coatings. The aim of this research, as a part of the Supergen Plant Life Extension (PLE) project, was (a) to investigate the influence of industrial gas turbine blade geometry on TBC system lifetimes, and (b) to extend knowledge on the effect of bond coat composition on the oxide growth at temperatures below 1000 °C. The main results of this thesis, obtained using mass change and characterisation techniques, increase the understanding of the significant interactions between the different coating layers, samples’ geometry, interdiffusion and failure mechanisms involved during oxidation. Curvature was found to affect the quality of manufacture and thus promoted premature failure at the convex features of modified aerofoil-shaped samples. In parallel new bond coat compositions, suitable for industrial gas turbines were suggested from the wide range investigated in oxidation exposures. The selective growth of protective Cr2O3 or Al2O3 oxides or other mixed oxides was observed and mapped in ternary diagrams. Furthermore two novel techniques were successfully used during this project. Pulsed flash thermography proved to be efficient in identifying areas of sub-surface TBC delamination non-destructively. Magnetron co-sputtering using 2 and 3 targets was found to be a flexible method to deposit thick coatings with a wide range of compositions.

A possible shaft failure event can trigger a number of mechanisms affecting the mechanical integrity particular of turbine discs and blades. A predominant aim in engine design and development is to satisfy that passenger lives are not to be endangered by the release of high energy debris. In a typical Intermediate Pressure shaft failure scenario of a 3-spool high bypass ratio turbofan engine, a potential mechanism to limit the terminal speed of the free running turbine, within acceptable values, is proven to be the impact of the free running turbine with the following stationary arrangement. Cont/d.

Accurate and reliable component life prediction is crucial to ensure both the safety and economics of gas turbine operations. In the pursuit of improved accuracy and reliability, current model-based creep life estimation methods have become more and more complicated and therefore demand huge amounts of work and significant amounts of computational time. Because of the underlying problems arising from current life estimation methods, this research aims to develop an alternative performance-based creep life estimation method that is able to provide a quick solution to creep life prediction while at the same time maintaining the achieved accuracy and reliability as that of the model-based method. Using an artificial neural network, the existing creep life prediction subprocesses and secondary inputs are ‘absorbed’ into simple parallel computing units that are able to create direct mapping between various gas turbine operating and health conditions or gas path sensors and creep life. The outcome of this research is the creation of three proposed neural-based creep life prediction architectures known as the Range-Based, Functional-Based and Sensor-Based. An integrated creep life estimation model was first developed and incorporated into an in-house performance simulation and diagnostics software. Using the integrated model, the effects of several operating and health parameters on a selected turbo-shaft engine model turbine blade’s creep life was initially performed using an introduced Creep Factor approach. The outcomes of this investigation were then used to populate input-output samples to train and validate the neural-based creep life prediction architectures. To ensure that the proposed neural architectures are able to achieve generalisation and produce accurate creep life prediction for both clean and degraded engine conditions, four-stage assessments were carried out. Finally, the effects of input uncertainties on the creep life prediction were investigated to assess how sensitive the proposed architectures are to different levels of uncertainty. The results show that all of the proposed neural architectures were able to produce accurate creep life predictions for both clean and degraded engine conditions. When comparing the three proposed architectures, the Sensor-Based architecture was found to be the most accurate in both conditions. Despite the accurate creep life prediction, it was also found that all of the proposed architectures were sensitive to input uncertainties with the Functional-Based architecture being the least sensitive to the uncertainty.

A gas turbine engine can over-speed due to various reasons, including shaft failure, variable geometry mal-schedule or fuel system malfunction. In any case, engine manufacturers are required to demonstrate that a shaft over-speed event will not result in an uncontained failure with high energy debris being released from the engine. Although the certification authority can be satisfied that the engine is shaft failure safe by conducting large scale tests, a purely experimental approach would be very complex and expensive. Moreover, today’s poor understanding of the event leads to conservative designs that exert unavoidable penalties on the engine performance and weight. It is in this context that the need for an analytical approach and small scale testing arises to model the progression of the event. This work is part of a wider long term research collaboration between Cranfield University and Rolls-Royce that attempts to enhance today’s modelling capability of gas turbine shaft overspeed/ failure events. The final aim of the project is the development of a generic advanced performance prediction tool able to account for all the complex and heavily interrelated phenomena to ascertain the terminal speed of the over-speeding turbine. This multidisciplinary “all in one” tool will allow to include the shaft failure scenario early into the design process and eliminate current conservative design approaches while maintaining the high standard of airworthiness required for certification. This thesis focuses on the aerothermal performance modelling of turbomachinery components at the extreme off-design conditions experienced during a shaft over-speed event. In particular, novel modelling techniques and methodologies at the forefront of knowledge have been developed to simulate the performance of turbine vanes at high negative incidence angles, derive the extended compressor characteristics in reverse flow and calculate the response of the air system during rapid transient among others. These component models are ready to be integrated into a generic single computational tool that, once validated against engine data available from the sponsor, can be applied to different engines and scenarios. The collaboration with Rolls-Royce provided the opportunity to conduct research on other areas related to performance engineering apart from the shaft failure modelling. The present study makes several noteworthy contributions on compressor variable geometry loss, deviation and stall modelling, compressor variable geometry schedule optimisation and on the effect of using real gas models instead of the perfect gas assumption in engine performance simulation codes.