No description available.

No description available.

No description available.

No description available.

There is a common requirement in the process, oil and gas turbine industries for high performance axial flow compressors operating over a wide range of mass flow rate and rotational speed at high efficiency. The trends have been for higher blade loadings (greater pressure rise per stage) and higher efficiency which are increasingly achieved through sophisticated Computational Fluid Dynamics designs. These trends, however, tend to mitigate against stable operating flow range (reduced surge margin), which can often lead to performance compromises. The objective of this work is to investigate the possibility of using alternative means to gain ow range by better use of variable geometry, which may permit design objectives to be better achieved. Variable geometry of the type envisaged is already often employed to overcome part-speed operating problems, but it proposed here that there may be additional benefits from their more intelligent control. The operation of axial compressors with a wide operating range is limited by instabilities, which cause a full breakdown of the flow, which is surge. These instabilities, which are caused by high incidence and subsequent stalling of stages occur due to different phenomena at part and full speed operation. The problem at part-speed is that the front stages are often heavily stalled and rear stages choked, whereas at high speeds, the front stages are operating close to choke and the rear stages tend to be stalling. Optimisation of the design to full load conditions can often provide part-speed problems and to achieve the acceptable performance, variable geometry over the front region of the compressor is sometimes used to modify the flow angles and avoid stage stall and subsequent surge. To-date, such variable settings follow some schedule established by analysis and experiment whereas this work presents a methodology of setting blade rows using an optimisation procedure and investigates the likelihood of performance benefits being obtained by a control technique which reacts° to these changing conditions. The construction of the numerical method presented in this thesis was done with an emphasis upon its intended contribution towards a eventual online control application. Therefore, a practical approach has been employed in the development of the compressor modelling techniques used in the work. Specifically, a highly empirical one-dimensional performance prediction code was constructed, employing successful correlations taken from the literature. This was coupled to a surge prediction method that has been shown in the past to function more than satisfactorily in a multistage environment. Finally, the predicted stage and overall performance (including the surge point) characteristics were passed to a optimisation program, which allowed these simulated conditions to be investigated. It is hoped that the work presented has illustrated the potential (from a aerodynamic performance point of view) of such a control technique to offer additional freedom in the operation of a multistage axial flow compressor. Moreover, the numerical modelling techniques have been developed enough to envisage (at least in part) their simple integration within a practical system. Clearly, some further investigations are required to take this work forward and the next logical step would be to improve the empirical rules with which the blade performance is predicted. A experimental programme would also be of great advantage, for example in the study of how the deviation angle for a given blade row varies over time (operating hours) in a real machine due to ageing and fouling. This would allow better estimates of the stage work during long term operation so that the optimiser could adapt to the slowly degrading performance of the blades. Finally, it is important to verify the simulated results with measured data, taken at the same optimal stator vane settings as given by the program. This must be carried out before it can be applied to a real application, although a limited study of this nature is presented in chapter 6.

No description available.

An efficient FAS muldgrid solution strategy is presented for the accurate and economic simulation of convection dominated flows. The use of a high-order approximation to the convective transport terms found in the governing equations of motion has been investigated in conjunction with an unsegregated smoothing technique. Results are presented for a sequence of problems of increasing complexity requiring that careful attention be directed toward; the proper treatment of different types of boundary condition. The classical two-dimensional problem of flow in a lid-driven cavity is investigated in depth for flows at Reynolds numbers of 100,400 and 1000. This gives an extremely good indication of the power of a multigrid approach. Next, the solution methodology is applied to flow in a three-dimensional lid-driven cavity at different Reynolds numbers, with cross-reference being made to predictions obtained in the corresponding two-dimensional simulations, and to the flow over a step discontinuity in the case of an abruptly expanding channel. Although, at first sight, these problems appear to require only minor extensions to the existing approach, it is found that they are rather more idiosyncratic. Finally, the governing equations and numerical algorithm are extended to encompass the treatment of thermally driven flows. Ile solution to two such problems is presented and compared with corresponding results obtained by traditional methods.

This thesis presents a detailed investigation of the secondary flow and boundary layers in a large scale, linear cascade of high pressure turbine rotor blades. The puropose of the data is to provide a suitable test case to aid the design and validation of the turbulence and transition models used in computational fluid dynamics. Hot-wire measurements have been made on a number of axial planes upstream, within and downstream of the blades to give both the mean flow conditions and all six components of Reynolds stress. Suitable inlet conditions have been defined at one axial chord upstream of the blade leading edge where the velocity and turbulence have been measured in both the freestream and endwall boundary layer. The turbulence dissipation rate has also been measured in order to define fully the inlet flow, a quantity that is usually missing in other data. Measurements through the blade show that the turbulence generation associated with the secondary flows is considerable and that all three shear stress components are significant. Intermittency measurements close to the endwall and blade surfaces show that the boundary layers are mostly laminar or transitional. The new endwall boundary layer, that forms behind the separation line, was found to be initially laminar. On the suction surface transition occurs over the latter part of the blade and on the pressure surface the accelerating flow causes relaminarisation. A number of calculations using a mixing length and high and low Reynolds number k-ϵ calculations show that reasonable overall results may be obtained. The lack, or failure, of transition modelling caused profile losses to be generally overpredicted and there was little evidence that the more sophisticated models produced better results. No model accurately predicted the individual turbulence quantities largely due to the inadequacy of the Boussinesq assumption for this type of flow. Good transition modelling appears to be more important than turbulence modelling in terms of the overall results.

No description available.

This thesis describes the development of a numerical solution procedure which is valid for both incompressible flow and compressible flow at any Mach number. Most of the available numerical methods are for incompressible flow or compressible flow only and density is usually chosen as a main dependent variable by almost all the methods developed for compressible flow. This practice limits the range of the applicability of these methods since density changes can be very small when Mach number is low. Even for high Mach number flows the existing time-dependent methods may be inefficient and costly when only the finial steady-state is of concern. The presently developed numerical solution procedure, which is based on the SIMPLE algorithm, solves the steady-state form of the Navier-stokes equations, and pressure is chosen as a main dependent variable since the pressure changes are always relatively larger than the density changes. This choice makes it possible that the same set of variables can be used for both incompressible and compressible flows. It is believed that Reynolds stress models would give better performance in some cases such as recirculating flow, highly swirling flow and so on where the widely used two equation k-e model performs poorly. Hence, a comparative study of a Reynolds stress model and the k-e model has been undertaken to assess their performance in the case of highly swirling flows in vortex throttles. At the same time the relative performance of different wall treatments is also presented. It is generally accepted that no boundary conditions should be specified at the outflow boundary when the outflow is supersonic, and all the variables can be obtained by extrapolation. However, it has been found that this established principle on the outflow boundary conditions is misleading, and at least one variable should be specified at the outflow boundary. It is also shown that the central differencing scheme should be used for the pressure gradient no matter whether it is subsonic or supersonic flow.