Approved for public release; distribution is unlimited / This thesis presents a simple, meanline analysis of the impact of blade roughness on the mass flow, work coefficient, and efficiency of a three-stage axial compressor as a function of the location of fouling. First, an extensive review is presented on the state-of-the-art of measuring compressor degradation and on the impact of roughness on loss and deviation in a compressor cascade. The performance of a baseline, three-stage compressor, which has hydrodynamically smooth blades, is predicted. Using this baseline geometry, the influence of roughness in the front, middle and rear stages is calculated using empirical data for the enhanced losses and increased deviation, with a stage stacking technique. Influence coefficients that relate percentage changes in one parameter to percentage changes in other parameters are calculated. This analysis predicts that the most sensitive parameter for predicting fouling in the front stages is the percentage change in mass flow and the most sensitive parameter for predicting fouling in the rear stages is the efficiency. / Lieutenant, United States Coast Guard

The high pressure turbine stage can be considered the most important component for the efficiency and longevity of a modern gas turbine. The flow field within this stage is highly complex and is both unsteady and three-dimensional. Understanding this flow field is essential if improvements are to be made to future engine designs. Increasingly designers are placing more emphasis on the use of Computational Fluid Dynamics (CFD) rather than experimental results. CFD methods can be more flexible and cost effective. However before these predictions can be used they must be validated against experimental data at engine conditions. The hostile environment and complexity of flows within a gas turbine engine mean that collection of experimental data is extremely challenging. This thesis describes the development of an instrumentation technique for unsteady gas turbine flow measurement capable of resolving unsteady three-dimensional flow. The technique is based on an aerodynamic probe constructed with miniature semiconductor pressure transducers manufactured by Kulite Semiconductor Inc. Measurements recorded using this instrumentation technique from the Oxford Rotor experiment are presented to illustrate its use, and these in turn are compared with a CFD prediction of the rotor flow-field. This work was funded by the Engineering and Physical Sciences Research Council and Kulite Semiconductor Inc. The Oxford Rotor project is jointly funded by the Engineering and Physical Sciences Research Council (EPSRC), and Rolls-Royce Plc.

The ternary Pt-Al-Cr system was investigated as part of the continued development of a thermodynamic database for the Pt-Al-Cr-Ru system. Scanning electron microscopy with energy dispersive X-ray spectroscopy and X-ray diffraction analyses were used to obtain phase equilibria data. The alloys were studied in the as-cast condition, as well as after annealing at 600°C and 1000°C respectively. A solidification projection was constructed and a liquidus surface derived. Isothermal sections at 600°C and 1000°C were also determined. It was concluded that all phase regions were identified correctly since the results were selfconsistent. Three ternary phases were found and 19 ternary invariant reactions identified. A thermodynamic database was developed for the Pt-Al-Cr system using Thermo-Calc. Phase relations could be reasonably accurately predicted between 600°C and 1000°C, and even up to temperatures close to the melting point. However, the match between the calculated and experimental diagrams could be improved. As with the Pt-Cr-Ru system, problems with the constituting binary systems seemed to be the major cause for problems encountered in the modelling. Only once the Al-Pt and especially the Cr-Pt and Cr-Ru binary phase diagrams are confirmed more rigorously, the calculated ternary phase diagrams could be worked on with more confidence. More than half of the alloys investigated had hardnesses in excess of 600 HV10 regardless of their state of heat treatment. Based on the examination of hardness indentations, alloys in the Pt-Al-Cr system were also often brittle due to the presence of hard intermetallic compounds. Alloys containing ~Pt3Al showed better behaviour with regard to toughness which was encouraging for the Pt-based alloys that are being developed by Mintek.

The effects of the orifice geometry and the surface boundary condition on the heat transfer distribution to a flat surface of an impinging jet array were investigated. The jet array impinged normally onto the surface which was either isothermal or had a uniform heat flux. The experiments were performed for the flow rate range from 0.0039 to 0.0070 m��/s corresponding to jet Reynolds numbers of 5000 to 11000. The jet-to-surface spacings varied from 1 to 4 jet diameters. After impinging, the air jet was constraine4 to exit in one direction creating a "crossflow". condition. The isothermal surface results are presented in terms of the average heat transfer coefficient. For the uniform heat flux surface, both average and local values are presented. The average and local heat transfer distributions were mapped using thermochromic liquid crystals. Results are presented for two jet geometries: circular and cusped ellipse. The cusped ellipse jets show better heat transfer performance compared to the circular jets for both surface boundary conditions. This is thought to be a result of increased turbulence and the axis-switching phenomenon. Results for the uniform heat flux surface boundary higher than for the isothermal surface boundary condition. This result can be explained by the difference between the surface temperature and the jet temperature for both surface boundary conditions. Correlations of Nusselt versus Reynolds numbers are presented for both jet geometries and surface boundary conditions. / Graduation date: 2000

Evaporative gas turbine cycles (EvGT) show an exceptionalexhaust heat recovery potential, which makes them a strongcompetitor to other advanced gas turbine cycles, especiallyfrom small to intermediate sizes. Evaporative gas turbines aredistinguished by humidifying the working fluid beforecombustion at temperatures below the boiling point of water;and the heat required for evaporation of water is partly takenout of the exhaust gas. Thus, humidification is a key operationin these cycles. This thesis investigates, both theoreticallyand experimentally, two alternative approaches tohumidification: the packed-bed humidification tower and thetubular humidifier. Both these equipments involvecountercurrent contact between water and the working fluid.Humidifier design criteria are developed and criticalparameters such as flooding, wetting rate and entrainment arediscussed. The experimental parts were carried out on thepacked-bed tower in the EvGT pilot plant, and on a tubularhumidifier test rig especially erected for this purpose. Thetheoretical models were confirmed by the experiments. The height of a transfer unit, necessary for designingpacked beds, was calculated for the packing employed in theEvGT pilot plant. It was found that the data provided by themanufacturer may be used with minor corrections. The tubular test rig operated satisfactorily delivering hothumid air. The theoretical models coincided well with theexperimental results, verifying the design criteria developedhere. The heat transfer calculations indicated that mostresistance to heat transfer is on the exhaust gas side. Thus, asurface extended tube (Sunrod) was used in the test rig. Itcould be concluded that the tubular humidifier is a strongalternative to the packed-bedtower, especially in smallhigh-pressure gas turbines. Furthermore, the importance of the non-ideality of theair-water vapor mixture in modeling evaporative cycles wasfirst highlighted in this work. Through applying realthermodynamic properties of air-water vapor mixtures in cyclecalculations, it was found that the compressed air contains ahigher amount of moisture than indicated by the ideal gasmixture model. This affects the design of the heat recoverysystem and cannot be neglected. <b>Key words:</b>evaporative gas turbine, indirect-fired gasturbine, humidification, packed bed, tubular humidifier,evaporator, saturator.

<p>Evaporative gas turbine cycles (EvGT) show an exceptionalexhaust heat recovery potential, which makes them a strongcompetitor to other advanced gas turbine cycles, especiallyfrom small to intermediate sizes. Evaporative gas turbines aredistinguished by humidifying the working fluid beforecombustion at temperatures below the boiling point of water;and the heat required for evaporation of water is partly takenout of the exhaust gas. Thus, humidification is a key operationin these cycles. This thesis investigates, both theoreticallyand experimentally, two alternative approaches tohumidification: the packed-bed humidification tower and thetubular humidifier. Both these equipments involvecountercurrent contact between water and the working fluid.Humidifier design criteria are developed and criticalparameters such as flooding, wetting rate and entrainment arediscussed. The experimental parts were carried out on thepacked-bed tower in the EvGT pilot plant, and on a tubularhumidifier test rig especially erected for this purpose. Thetheoretical models were confirmed by the experiments.</p><p>The height of a transfer unit, necessary for designingpacked beds, was calculated for the packing employed in theEvGT pilot plant. It was found that the data provided by themanufacturer may be used with minor corrections.</p><p>The tubular test rig operated satisfactorily delivering hothumid air. The theoretical models coincided well with theexperimental results, verifying the design criteria developedhere. The heat transfer calculations indicated that mostresistance to heat transfer is on the exhaust gas side. Thus, asurface extended tube (Sunrod) was used in the test rig. Itcould be concluded that the tubular humidifier is a strongalternative to the packed-bedtower, especially in smallhigh-pressure gas turbines.</p><p>Furthermore, the importance of the non-ideality of theair-water vapor mixture in modeling evaporative cycles wasfirst highlighted in this work. Through applying realthermodynamic properties of air-water vapor mixtures in cyclecalculations, it was found that the compressed air contains ahigher amount of moisture than indicated by the ideal gasmixture model. This affects the design of the heat recoverysystem and cannot be neglected.</p><p><b>Key words:</b>evaporative gas turbine, indirect-fired gasturbine, humidification, packed bed, tubular humidifier,evaporator, saturator.</p>

Thesis (Ph. D.)--West Virginia University, 2003. / Title from document title page. Document formatted into pages; contains xiv, 185 p. : ill. (some col.). Vita. Includes abstract. Includes bibliographical references (p. 170-176).

Not available / text

In a previous work of a similar nature, the performance of a low speed axial turbine with a second stage nozzle was examined with respect to the effect of the variation of tip clearance for various tip shapes. Present findings suggest some interesting phenomena, including the effect of tip clearance on the flow within the rotor and show that poor resolution from a transducer and insufficient data points in the critical tip region, where a high velocity peak was found, were responsible for a number of incorrect conclusions in the original study. In terms of blade tip geometry, a standard flat tip shape was found to deliver only a marginally better performance when compared to a double squealer tip and the two streamlined shapes previously investigated. Although contemporary opinion suggests that a streamlined tip should increase the leakage flow and hence cause greater mixing losses, the machine efficiency was not significantly reduced. This is an exciting result since it suggests that a streamlined tip shape can be used to alleviate the problem of blade tip burnout without significantly reducing machine efficiency. When the single stage performance in the absence of a second nozzle was examined, slightly different trends were obtained. The low entropy tips produced slightly lower mixing loss, suggesting that the internal gap loss is an important parameter in determining the rate at which the leakage jet mixes downstream of the rotor. The flow behind the rotor (ie time averaged) was found to be in remarkable agreement with linear cascade data when time averaged even though the latter did not include any effects of relative motion. An increase in clearance was seen to reduce the Euler work and also to cause a deficit of mass flow across the remainder of the blade right down to the hub. The leakage flow was also seen to induce a flow blockage which resulted in a higher driving pressure across the rotor for the same mass flow rate. As in the previous study, the second stage nozzle efficiency was seen to be independent of tip clearance or tip shape and was moderately better than that of the first nozzle. However, the improvement was not found to be as large, due to a previously undetected very thin ring of high energy leakage fluid. When this is taken into account, the efficiency of the second stage nozzle is comparable to the first. The second nozzle was seen to have a flow straightening effect on the poorly deflected, high energy leakage flow, causing a rapid mixing process within these downstream blade passages. The growth of secondary flow was reduced at both the hub and the tip and this is believed to result in a slight decrease in loss. The outlet flow was closer to design conditions than that of the first stage nozzle. / Thesis (Ph.D.)-University of Natal, Durban, 1996.

With strict environmental regulations, gas turbine emissions have been heavily constrained. This requires operating conditions wherein thermo-acoustic flame instabilities are prevalent. During this process the combustor acoustics and combustion heat release fluctuations are coupled and can cause severe structural damage to engine components, reduced operability, and inefficiency that eventually increase emissions. In order to develop an engine without these problems, there needs to be a better understanding of the physics behind the coupling mechanisms of this instability. Among the several coupling mechanisms, the “velocity coupling” process is the main focus of this thesis. The majority of literature has treated axisymmetric disturbance fields which are typical of longitudinal acoustic forcing and axisymmetric excitation of ring vortices. Two important non-axisymmetric disturbances are: (1) transverse acoustics, in the case of circumferential modes of a multi-nozzle annular combustor and (2) helical flow disturbances, seen in the case of swirling flow hydrodynamic instabilities. With significantly less analytical treatment of this non-axisymmetric problem, a general framework is developed for three-dimensional swirl-stabilized flame response to non-axisymmetric disturbances. The dynamics are tracked using a level-set based G-equation applicable to infinitely thin flame sheets. For specific assumptions in a linear framework, general solution characteristics are obtained. The results are presented separately for axisymmetric and non-axisymmetric mean flames. The unsteady heat release process leads to an unsteady volume generation at the flame front due to the expansion of gases. This unsteady volume generation leads to sound generation by the flame as a distributed monopole source. A sound generation model is developed where ambient pressure fluctuations are generated by this distributed fluctuating heat release source on the flame surface. The flame response framework is used to provide this local heat release source input. This study has been specifically performed for the helical flow disturbance cases to illustrate the effects different modes have on the generated sound. Results show that the effects on global heat release and sound generation are significantly different. Finally, the prediction from the analytical models is compared with experimental data. First, a two-dimensional bluff-body stabilized flame experiment is used to obtain measurements of both the flow and flame position in time. This enables a local flame response comparison since the data are spatially resolved along the flame. Next, a three-dimensional swirl-stabilized lifted flame experiment is considered. The measured flow data is used as input to the G-equation model and the global flame response is predicted. This is then compared with the corresponding value obtained using global CH* chemilumenescence measurements.