Design optimization method for axial turbines.

摘要

Designing axial flow turbines is a complex process requiring significant time and effort in the earliest phases. It is necessary to conduct numerous nested iterations prior to arriving at a stage with potential for high efficiency performance. The process is confounded by limitations in methods available in the literature where key design parameters like rotor diameter or efficiency are treated as dependant rather than independent variables, preventing the design from being optimized. This thesis presents a new method for designing high efficiency axial flow turbines. This new method significantly reduces the complexity of design effort and time, enabling cost effective design optimization.; Multivariate regression analysis is used to produce an equation relating the efficiency to the design parameters. A standard solver package is used to identify the specific values of application constrained design parameters, which yield optimum efficiency. A software tool, Turbine Design Analysis Tool (TDAT), outputs a dataset completely describing the turbine stage flow passages in three dimensions and is in a format readily imported into CFD and FEA software for more advanced analysis occurring in later phases of the design process.; The results were verified using empirical data and computational fluid dynamics. It was concluded that the new method significantly reduces design complexity and time. However future research is recommended to determine the feasibility of achieving a better fit to the data that would improve the quality of the new method.