No description available.

No description available.

No description available.

No description available.

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.

No description available.

The successful design of a long-lived and efficient gas turbine engine requires a good knowledge of the thermal and aerodynamic performances of the components of the turbine. Of particular importance, is the heat transfer rate from the hot gases to the cooled turbine blades, since this limits the maximum turbine entry temperatures which can be obtained. Much gas turbine research is concentrated on experimental modelling and measurements to assist in the development of improved theoretical prediction techniques. The difficulties of instrumenting fully rotational rigs, which are necessary for a full understanding of the complex three dimensional flow in the turbine, have, however, to a large extent, limited most experimental research to stationary facilities. A technique is described which will allow heat transfer rate measurements to be made on fully rotating test facilities using mutlilayered model turbine blades comprising an electrical insulator on a metal base. An accurate and computationally efficient method for determining the surface heat flux to a multi-layered model turbine blade is developed theoretically, together with a method for calibrating the thermal properties of the multi-layered system. This method allows the existing successful heat flux measurement technique, which utilises electronic analogue circuitry in conjunction with thin film surface thermometers on a model made from a thermal insulator, to be extended for application to multi-layered models. The production of test models by the application of a vitreous enamel (as an electrical insulator), to a mild steel, is identified as the most suitable coating technique for experimental application. Radiant and wind tunnel testing of multi-layered cylindrical models are described, which confirm that the method is both practical and accurate.

Film cooling by means of holes is an essential cooling technique in modern gas turbine engines. This cooling technique is employed over endwalls, as well as on the surface of blades. Thus, there is a need for film cooling predictions in a three-dimensional setting. Currently only boundary layer codes are available for design purposes and they are difficult to apply to the three-dimensional case with secondary flows. Present advanced computation prediction methods are capable of solving the complete flow field in three dimensions with coolant flow. However, the spatial resolution that these methods require eliminate them as suitable options for design tools This study introduces a simpler description of the film cooling process which may be implemented in a code for design purposes. The parameters of turbulence enhancement, turbulence decay, and the coolant distribution at injection were optimized using existing experimental data. Finally, the code was employed in a three-dimensional setting with film cooling present. An experimental study of the flow through cooling holes was also undertaken. Two unique geometries were later developed where a row of cooling holes exited into a vortex region where the flow was mixed before being injected from a slot. The cooling benefits of these geometries is apparent.

In this thesis the author presents two areas of work; exploring the integration of Computational Fluid Dynamics (CFD) into engine design for Jaguar Cars Ltd and developing a novel 'mesh construction' method for making mesh generation both easy and fast. It is concluded that Jaguar can use CFD in the evaluation stage of the engine design process, although not in the concept stage of design. The CFD predictions are shown to be useful for detecting flow related faults and determining the general flow trends, but they should not be used as an absolute measure of the flow variables. The author has determined an efficient method for obtaining good quality meshes using commercial modelling and mesh generation software which requires a skilled CFD analyst. Steady flow analysis of an engine port and cylinder design could currently be completed in about six weeks using a high-powered workstation. The author recommends dedicated workstations for CFD analysis and training Jaguar's draughtsmen to create CAD models with computer analysis requirements in mind. The author's mesh construction program automatically joins two overlapping meshes or cuts one mesh from another. Whilst the program works well on the test cases considered, it is not at a stage for commercial exploitation. Further development is therefore recommended.

No description available.